Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus)...

91
Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll, JE Feenstra, WB Jackson and MT Lloyd SARDI Publication No. F2007/000523-4 SARDI Research Report Series No. 930 SARDI Aquatic Sciences PO Box 120 Henley Beach SA 5022 October 2016 Fishery Assessment Report to PIRSA Fisheries and Aquaculture

Transcript of Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus)...

Page 1: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

i

Snapper (Chrysophrys auratus) Fishery

AJ Fowler, R McGarvey, J Carroll, JE Feenstra, WB Jackson and MT Lloyd

SARDI Publication No. F2007/000523-4 SARDI Research Report Series No. 930

SARDI Aquatic Sciences

PO Box 120 Henley Beach SA 5022

October 2016

Fishery Assessment Report to PIRSA Fisheries and Aquaculture

Page 2: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

ii

Snapper (Chrysophrys auratus) Fishery

Fishery Assessment Report to PIRSA Fisheries and Aquaculture

AJ Fowler, R McGarvey, J Carroll, JE Feenstra, WB Jackson and MT Lloyd

SARDI Publication No. F2007/000523-4 SARDI Research Report Series No. 930

October 2016

Page 3: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

iii

This publication may be cited as: Fowler, A.J., McGarvey, R., Carroll, J., Feenstra, J.E., Jackson, W.B. and Lloyd, M.T. (2016). Snapper (Chrysophrys auratus) Fishery. Fishery Assessment Report to PIRSA Fisheries and Aquaculture. South Australian Research and Development Institute (Aquatic Sciences), Adelaide. F2007/000523-4. SARDI Research Report Series No. 930. 82pp.

South Australian Research and Development Institute SARDI Aquatic Sciences 2 Hamra Avenue West Beach SA 5024

Telephone: (08) 8207 5400 Facsimile: (08) 8207 5406 http://www.pir.sa.gov.au/research DISCLAIMER The authors warrant that they have taken all reasonable care in producing this report. The report has been through the SARDI internal review process, and has been formally approved for release by the Research Chief, Aquatic Sciences. Although all reasonable efforts have been made to ensure quality, SARDI does not warrant that the information in this report is free from errors or omissions. SARDI does not accept any liability for the contents of this report or for any consequences arising from its use or any reliance placed upon it. The SARDI Report Series is an Administrative Report Series which has not been reviewed outside the department and is not considered peer-reviewed literature. Material presented in these Administrative Reports may later be published in formal peer-reviewed scientific literature. © 2016 SARDI This work is copyright. Apart from any use as permitted under the Copyright Act 1968 (Cth), no part may be reproduced by any process, electronic or otherwise, without the specific written permission of the copyright owner. Neither may information be stored electronically in any form whatsoever without such permission. Printed in Adelaide: October 2016 SARDI Publication No. F2007/000523-4 SARDI Research Report Series No. 930 Author(s): AJ Fowler, R McGarvey, J Carroll, JE Feenstra, WB Jackson and

MT Lloyd Reviewer(s): M Steer, G Grammer (SARDI) and J McPhail (PIRSA) Approved by: S Mayfield Science Leader - Fisheries Signed: Date: 24 October 2016 Distribution: PIRSA Fisheries and Aquaculture, SAASC Library, SARDI Waite

Executive Library, Parliamentary Library, State Library and National Library

Circulation: Public Domain

Page 4: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

iv

TABLE OF CONTENTS

TABLE OF CONTENTS ........................................................................................... iv

LIST OF TABLES ..................................................................................................... v

LIST OF FIGURES ................................................................................................... vi

ACKNOWLEDGEMENTS ........................................................................................ ix

EXECUTIVE SUMMARY........................................................................................... 1

1. INTRODUCTION .............................................................................................. 3 1.1 Overview ........................................................................................................ 3 1.2 Description of the fishery ................................................................................ 3 1.3 Biology of snapper ......................................................................................... 5 1.4 Research program ......................................................................................... 9 1.5 Information sources for stock assessment ..................................................... 9 1.6 Harvest strategy ........................................................................................... 11 1.7 Stock status classification ............................................................................ 14

2. METHODS ..................................................................................................... 15 2.1 Commercial fishery statistics ........................................................................ 15 2.2 Recreational fishery statistics ....................................................................... 15 2.3 Market sampling .......................................................................................... 16 2.4 Stock assessment model – SnapEst ............................................................ 17 2.5 Assessment of fishery performance ............................................................. 18

3. RESULTS ...................................................................................................... 20 3.1 Commercial fishery ...................................................................................... 20 3.2 Recreational fishery ..................................................................................... 35 3.3 Stock assessment model - SnapEst ............................................................. 36 3.4 Year class contributions to fishery catches................................................... 41 3.5 Fishery performance .................................................................................... 42

4. DISCUSSION ................................................................................................. 45 4.1 Information for stock assessment ................................................................. 45 4.2 Status of snapper stocks .............................................................................. 46 4.3 Conclusions ................................................................................................. 52 4.4 Uncertainties in the assessment .................................................................. 53 4.5 Future research needs ................................................................................. 54

5. REFERENCE LIST ........................................................................................ 56

6. APPENDICES ................................................................................................ 60 6.1 Annual size and age structures .................................................................... 60 6.2 Snapper stock assessment model ............................................................... 71 6.3. Estimating length-at-age and weight-length for the two newly implemented Gulf St. Vincent regions .............................................................................................. 79

Page 5: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

v

LIST OF TABLES

Table 1.1 Fishery performance indicators and trigger reference points used to assess fishery performance as specified in the Management Plan (PIRSA 2013). The type of indicator and whether a primary or secondary one is also indicated. G – general; B – biological; P – primary; S – secondary. ......................................................................................................................... 13

Table 1.2 Allocation percentages and trigger limits for SA’s snapper fishery. Table shows the percentage of total allocation amongst commercial and recreational sectors, and trigger reference point for each sector (Trigger 1). It also shows commercial allocation and trigger reference points for Triggers 2 and 3. Fishing sectors are; MSF = Marine Scalefish, SZRL = Southern Zone Rock Lobster, NZRL = Northern Zone Rock Lobster, LCF = Lakes and Coorong, REC = general recreational, CHTR = charter boat, ABT = aboriginal/ traditional. .................. 13

Table 1.3 Classification scheme used to ascribe fishery stock status. The description of each stock status and its potential implications for fishery management are also shown. .............. 14

Table 3.1 Comparisons of percentages of total catch of snapper taken by the different fishery sectors with their allocations and trigger limits specified in the Management Plan (PIRSA 2013). Trigger 1 relates to commercial and recreational sectors. Triggers 2 & 3 relate only to commercial fishery data. MSF – Marine Scalefish, SZRL – Southern Zone Rock Lobster, NZRL – Northern Zone Rock Lobster, LCF – Lakes and Coorong. ................................................... 42

Table 3.2 Fishery performance indicators and trigger reference points used to assess fishery performance as identified in the Management Plan (PIRSA 2013). The type of indicator and whether a primary or secondary one is also indicated. G – general; B – biological; P – primary; S – secondary. ......................................................................................................................... 44

Page 6: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

vi

LIST OF FIGURES

Fig. 1.1 Map of the coast of south eastern Australia, showing the stock structure for snapper based on fish movement (Fowler 2016). The arrows indicate directions and extent of emigration of fish from three primary nursery areas in Northern Spencer Gulf, Northern Gulf St. Vincent and Port Phillip Bay, Victoria. Inset shows the broader geographic region. SG – Spencer Gulf, GSV – Gulf St. Vincent, WC – west coast of Eyre Peninsula. ........................... 7

Fig. 2.1 Map of South Australia showing the six regional management units that are considered in the stock assessment processes. Inset shows the location of the geographic region along Australia’s southern coastline. NSG – Northern Spencer Gulf, SSG – Southern Spencer Gulf, NGSV – Northern Gulf St. Vincent, SGSV – Southern Gulf St. Vincent, SE – South East, WC – West Coast. .......................................................................................................................... 17

Fig. 3.1 State-wide annual commercial fishery statistics by calendar year from 1984 to 2015. a. catches of snapper by commercial fishery and gear type. b. estimates of targeted fishing effort by fishery and gear type. c. numbers of license holders who reported taking and targeting snapper in each year. .............................................................................................................. 21

Fig. 3.2 Relative estimates of total annual commercial catch of snapper by region, indicated by sizes of circles. ........................................................................................................................ 22

Fig. 3.3 Northern Spencer Gulf. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1). ...................................................... 24

Fig. 3.4 Southern Spencer Gulf. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1). ...................................................... 26

Fig. 3.5 Northern Gulf St. Vincent. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1). ...................................................... 28

Fig. 3.6 Southern Gulf St. Vincent. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential

Page 7: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

vii

data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1). ...................................................... 30

Fig. 3.7 South East. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1). ......................................................................................... 32

Fig. 3.8 West Coast. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1). ......................................................................................... 34

Fig. 3.9 Recreational fishery statistics from three State-wide telephone diary surveys. ......... 35

Fig. 3.10 Time series of the four annual biological performance indicators from the SnapEst fishery assessment model for Northern Spencer Gulf. Green lines show averages over the last three years, compared with blue lines giving averages over the three preceding years for biomass and six preceding years for recruitment, as in the trigger reference points for these indicators in Table 3.2.............................................................................................................. 37

Fig. 3.11 Time series of the four annual snapper biological performance indicators from the SnapEst model for Southern Spencer Gulf. ............................................................................ 38

Fig. 3.12 Time series of the four annual snapper biological performance indicators from the SnapEst model for Northern Gulf St. Vincent. ......................................................................... 39

Fig. 3.13 Time series of the four annual snapper biological performance indicators from the SnapEst model for Southern Gulf St. Vincent. ........................................................................ 40

Fig. 3.14 Estimates of annual total catches for each region and percentage of population that recruited prior to 2000, determined from annual age structures. ............................................ 41

Fig. 6.1 Size and biomass distributions for snapper caught in NSG in each of 13 years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class. ............................................................................................ 61

Fig. 6.2 Estimated age structures of fish caught in NSG in each of 13 years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned. ............................................................................. 62

Fig. 6.3 Size and biomass distributions for snapper caught in SSG in each of 12 years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class. ............................................................................................ 63

Page 8: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

viii

Fig. 6.4 Estimated age structures of fish caught in SSG in each of 12 years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned. ............................................................................. 64

Fig. 6.5 Size and biomass distributions for snapper caught in NGSV in each of eight years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class. ............................................................................................ 65

Fig. 6.6 Estimated age structures of fish caught in NGSV in each of seven years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned. ................................................................... 66

Fig. 6.7 Size and biomass distributions for snapper caught in SGSV in each of 12 years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class. ............................................................................................ 67

Fig. 6.8 Estimated age structures of fish caught in SGSV in each of seven years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned. ................................................................... 68

Fig. 6.9 Size and biomass distributions for snapper caught in SE in each of seven years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class. ............................................................................................ 69

Fig. 6.10 Estimated age structures of fish caught in SE in each of six years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned. ............................................................................. 70

Fig. 6.11 Fitted ogives of growth as length-at-age (graphs a and b, for Northern Gulf St. Vincent and Southern Gulf St. Vincent regions respectively) and weight-at-length (c and d). See text for details. Solid blue lines plot the fitted means (Eq. 2 for a and b, Eq. 6 for c and d). Blue error

bars show estimated 95% confidence intervals (1.96 times estimated ( )ia for a and b, and

1.96 times ( )w il for c and d) surrounding all fitted data points shown. ................................ 82

Page 9: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

ix

ACKNOWLEDGEMENTS

We wish to extend our gratitude to the workers at Adelaide’s SAFCOL fish market,

who provided us with ready access to the snapper catches at the market and a place

to work. Furthermore, we thank the numerous fish buyers who allowed us to dissect

their purchased fish. We thank Mike Steer and other workers from SARDI for their

significant contribution to the market sampling program. Overall the data collected

from market sampling has made an invaluable contribution to our understanding of

the biology of snapper.

The data on catch and effort from the commercial sector of the Marine Scalefish

Fishery were provided by Angelo Tsolos and Melleesa Boyle of the Information

Systems and Database Support Program of SARDI (Aquatic Sciences). The report

was reviewed by both Dr Michael Steer (SARDI) and Dr Gretchen Grammar

(SARDI), whose comments helped improve its content and presentation.

Page 10: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

1

EXECUTIVE SUMMARY

Jurisdiction South Australia

Stock SG/WC GSV WVS

Stock status Transitional–

depleting Sustainable Sustainable

Indicators Catch, CPUE, age structures, biomass

Catch, CPUE, age structures, biomass

Catch, CPUE, age structures,

Recruitment surveys

This is the 11th stock assessment report for the South Australian snapper fishery since

1997 and comes three years after the last assessment that was undertaken in 2013.

This assessment follows significant management changes implemented in 2012 and

2013, in response to concerns about fishery sustainability in several regions.

This assessment was done in the context of a recently completed biological study that

determined that the six South Australian regional populations involve three stocks. The

Spencer Gulf / West Coast Stock (SG/WC) involves Northern and Southern Spencer

Gulf (NSG, SSG) and the west coast of Eyre Peninsula (WC). The Gulf St. Vincent

Stock (GSV) includes Northern and Southern Gulf St. Vincent (NGSV, SGSV). The

South East region (SE) is part of the Western Victorian Stock (WVS). Fishery

performance indicators were assessed at the regional population scale, but status was

assigned at the scale of stock.

Commercial fishery statistics provided the ‘general’ fishery performance indicators and

two specific indicators for snapper, i.e. the proportions of daily catches that exceeded

250 kg for the handline and longline sectors (Prop250kgHL, Prop250kgLL). Also,

‘biological’ performance indicators included population age structures, as well as time-

series of estimates of fishable biomass, recruitment, exploitation rate, and egg

production that were output from the fishery assessment model SnapEst. For all

indicators, the estimate for 2015 or an average of the last three years was assessed

against reference points calculated from the historical time series from 1984 to 2015.

For NSG and SSG from 2012 to 2015, total catch, as well as targeted catch, effort and

catch per unit effort (CPUE) levels were all low, some at their historical lowest. For the

WC, commercial fishery statistics had declined since 2008. These regional declines

Page 11: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

2

point to low or declining levels of biomass for the SG/WC Stock. The age structures for

NSG and SSG indicated that recruitment through the 2000s had been poor relative to

strong year classes in the late 1990s. In contrast, the SnapEst model estimates for

both NSG and SSG indicated increasing or relatively high levels of biomass, in

response to low exploitation rates and the low and declining levels of commercial

fishing effort.

For NGSV, fishery statistics in 2015 remained at near record high levels, consistent with

high biomass, relating to a succession of strong year classes having recruited through

the 2000s. Model outputs from SnapEst concurred with respect to recruitment and

indicated that fishable biomass was at a record level, although also experiencing high

exploitation rates reflecting the high fishing effort. For SGSV, the fishery statistics were

relatively flat, following marginal declines since 2011. The recent age structures

demonstrated a number of strong year classes through the 2000s. These were

reflected in model outputs, which indicated an increasing biomass.

For the SE in 2015, record catches in recent years had declined, consistent with the

depletion of the recent, episodic high biomass level. The episodic fishery related to

strong 2001 and 2004 year classes having recruited to the population.

A total of 10 trigger reference points were breached for the general performance

indicators for the six regions. Breaches for NSG, SSG, SGSV, SE and WC related to

low estimates of fishery statistics. The one breach for NGSV related to high fishing

effort. There were also eight breaches for Prop250kgHL and Prop250kgLL, which

showed positive increases in the two Spencer Gulf regions, and a decrease for NGSV.

For the SG/WC Stock, substantial declines in catch, effort and CPUE suggest declining

levels of biomass, consistent with poor recruitment throughout the 2000s. Whilst in a

poor state, nevertheless with some recent recruitment, low recent fishing effort and with

management changes implemented in 2012 and 2013, this stock is classified as

transitional depleting. NGSV and SGSV had consistent fishery statistics, a

consequence of relatively high recruitment through the 2000s, having maintained high

biomass. As such, the GSV Stock is classified as sustainable. The SE region is part

of the WVS that has experienced relatively high recruitment over the past 12 years, and

will likely be replenished by emigration in the next few years. This stock is classified as

sustainable, although the regional population in the SE has produced considerably

lower catches than in recent years.

Page 12: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

3

1. INTRODUCTION

1.1 Overview

This report continues the triennial series of stock assessment reports for the South

Australian snapper (Chrysophrys auratus) fishery that have been done since 1997

(McGlennon and Jones 1997; McGlennon 1999; McGlennon and Jones 1999; Fowler

2000, 2002; Fowler et al. 2003, 2005, 2007, 2010, 2013). The aims of these reports

have been: to summarise information about the fishery and biology of the species; and

to use this information to determine the current status of the snapper stocks. The

previous stock assessment was completed in September 2013 (Fowler et al. 2013), and

summarised fishery and biological data up to 31 December 2012. This is the 11th report

in the series and incorporates data up to the end of December 2015.

1.2 Description of the fishery

1.2.1 Access

Snapper is an iconic fish species that supports important commercial and recreational

fisheries in each of the mainland States of Australia (Kailola et al. 2003). It is unusual

for an Australian marine fish species to have such a broad geographic distribution and

economic significance, which makes snapper one of Australia’s most significant fishery

resources. In recent years, South Australia (SA) has been one of the dominant State-

based contributors to the national total catches of both fishery sectors (Fowler et al.

2013). The snapper fishery is geographically extensive and encompasses most of the

State’s coastal marine waters from the far west coast to the south eastern region,

although the highest abundances and fishery catches are generally taken from Spencer

Gulf or Gulf St. Vincent.

Snapper is a primary target species of commercial and recreational industries of SA.

For the commercial industry, license holders from four different commercial fisheries

have access to SA’s snapper stocks, i.e. the Marine Scalefish Fishery (MSF), the

Northern Zone and Southern Zone Rock Lobster Fisheries (NZRLF, SZRLF) and the

Lakes and Coorong Fishery (LCF) (PIRSA 2013). Their main fishing gear types for

targeting snapper are handlines and longlines, since the use of hauling nets for taking

snapper was prohibited in 1993. Snapper is an icon species for local and inter-state

recreational fishers, with many particularly interested in the large trophy fish that can be

abundant in the coastal waters of SA. The recreational sector is divided into the

general and charter boat sectors. The former target snapper using rods and lines,

Page 13: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

4

primarily from boats, although jetty and land-based catches occur. The charter boat

sector is a commercial platform for recreational fishing that increases the probability of

successful operations, based on the expertise and equipment of licensed operators.

Recently, the commercial sector accounted for about 81% of SA’s snapper catch

(Jones 2009, PIRSA 2013). The 18% of total catch attributable to the recreational

sector involved 8% from the general recreational sector and 10% from the charter boat

sector (PIRSA 2013).

1.2.2 Management Regulations

Regulations for the commercial sector of South Australia’s snapper fishery involve a

suite of input and output controls (PIRSA 2013). The four commercial fisheries with

access to snapper each have limited-entry, which means that the numbers of fishers

who can target snapper have been limited for numerous years. There is a legal

minimum length of 38 cm total length (TL), whilst there are also several gear

restrictions. Snapper cannot be taken with fish traps, whilst the use of all nets,

including hauling nets and large mesh gill nets for targeting snapper was prohibited in

1993. From December 2012, the number of hooks on set lines was reduced from 400

to 200 for fishers operating within Spencer Gulf and Gulf St. Vincent, but remains at 400

for other regions. Also, a daily commercial catch limit of 500 kg was introduced for all

South Australian waters. There is also a 50 kg bycatch trip limit for the Commonwealth-

managed Southern and Eastern Scalefish and Shark Fishery. Commercial handline

fishers are limited with respect to the numbers of handlines and hooks per line that can

be legitimately used.

For the recreational sector, the minimum legal length of 38 cm TL applies, whilst bag

and boat limits vary geographically. In Gulf St. Vincent, Backstairs Passage and

Investigator Strait the bag limit is 5 and boat limit 15 fish for the size range of 38 - 60 cm

TL, whilst for all other State waters the bag limit is 10 and boat limit 30 fish. For fish

>60 cm TL, there is a bag limit of 2 fish and boat limit of 6 fish for all State waters. A

recent review of SA’s recreational fishery has recommended the removal of the

geographic variation in bag and boat limits (PIRSA 2016a), which would mean that the

bag limit of 5 fish and boat limit 15 fish for the size range of 38 - 60 cm TL would apply

for all State waters. For the Charter Boat sector the bag and boat limits are currently

similar to those of the general recreational sector, with a further complication relating to

number of passengers on board. When the number exceeds six passengers, the

Page 14: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

5

catches are limited to 1 fish per day for those >60 cm TL, and either 3 or 5 for the 38 –

60 cm TL, depending on location.

Since 2000, the management regime for snapper has involved at least one seasonal

closure per year which has applied for both fishing sectors. From 2003 to 2011, this

was a month-long fishery closure throughout November. From 2012, the seasonal

closure was extended for several weeks until 15th December for all fishing sectors.

Furthermore, in 2013, five snapper spawning spatial closures were implemented in the

northern gulfs to extend the duration of protection of important spawning aggregations

until the 31st January, thereby conferring protection for most of the reproductive season.

The four spatial closures in NSG and one in NGSV are fixed in location and circular with

a 4-km radius from a fixed point.

1.3 Biology of snapper

1.3.1 Distribution and stock structure

Snapper (Chrysophrys auratus) is a large, long-lived, demersal finfish species that is a

member of the family Sparidae (Perciforms: Percoidei). The species is broadly

distributed throughout the Indo-Pacific region including Japan, the Philippines, India,

Indonesia, New Zealand and Australia (Kailola et al. 1993). It has a broad Australian

distribution that includes the coastal waters of the southern two thirds of the continent,

extending southwards from the mid-coast of Western Australia, along the southern

continental coastline, and up the east coast as far as north Queensland (Kailola et al.

1993, Jackson et al. 2012). Snapper are also found along the north coast of Tasmania

and in recent years have extended their range to southern Tasmania (Last et al. 2011).

Throughout their broad distribution, snapper occupy a diversity of habitats from shallow,

demersal areas to the edge of the continental shelf across a depth range of 1 – 200 m.

Previous genetic studies have indicated that the Australian distribution of snapper is

divided into several separate stocks. One division is thought to occur at Wilson’s

Promontory in Victoria from where the East Coast Stock extends 2000 km up the east

coast as far as north Queensland (Sanders 1974). The Western Victorian Stock was

suggested to extend westwards from Wilson’s Promontory into south eastern South

Australia (SA), and to include the important nursery area of Port Phillip Bay (Donnellan

and McGlennon 1996). This was based on the analysis of allozymes, mitochondrial

DNA and tagging data which indicated a broad stock division between the Western

Victorian and the South Australian Stocks in the vicinity of the mouth of the Murray

Page 15: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

6

River. For the South Australian Stock that is distributed across >2000 km of

heterogeneous coastline and a diversity of habitats throughout the waters of Gulf St.

Vincent (GSV), Investigator Strait (IS), Spencer Gulf (SG), and the Great Australian

Bight (GAB), there is no evidence for any finer-scale genetic differentiation (Donnellan

and McGlennon 1996).

A recent study provided greater resolution to the understanding of stock structure for

snapper in south eastern Australia (Fowler 2016). It indicated, based on demographic

characteristics and the physical and chemical characteristics of otoliths, that South

Australian snapper involves three different stocks: the Spencer Gulf / West Coast Stock

(SG/WC); the Gulf St. Vincent Stock (GSV); and the Western Victorian Stock (Fowler

2016) (Fig.1.1). Each stock depends on a significant primary nursery area and

subsequent emigration of sub-adult and adult fish from that region to supplement

populations in adjacent regions. The latter stock extends westward from Port Phillip

Bay (PPB), Victoria into the south east region of SA, as far west as the Fleurieu

Peninsula. This stock depends on recruitment to PPB and subsequent westward

emigration of fish from that nursery area. As such, the population dynamics of this

episodic fishery are driven by inter-annual variability in recruitment to PPB. The

SG/WC Stock involves the regional populations of Northern Spencer Gulf (NSG),

Southern Spencer Gulf (SSG) and WC. NSG supports the source population and

nursery area from which fish emigrate to ultimately replenish SSG and the WC.

Northern Gulf St. Vincent (NGSV) supports the third primary nursery area and a self-

sustaining population. The area of SGSV and IS is likely to be a mixing zone for fish

from the three stocks.

Page 16: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

7

Fig. 1.1 Map of the coast of south eastern Australia, showing the stock structure for snapper based on fish movement (Fowler 2016). The arrows indicate directions and extent of emigration of fish from three primary nursery areas in Northern Spencer Gulf, Northern Gulf St. Vincent and Port Phillip Bay, Victoria. Inset shows the broader geographic region. SG – Spencer Gulf, GSV – Gulf St. Vincent, WC – west coast of Eyre Peninsula.

1.3.2 Growth, Longevity and Age Structures

Transverse sections of the sagittae of adult snapper from SA generally display clear

increments of opaque and translucent zones, which form annually and therefore relate

to fish age (McGlennon et al. 2000). Such sectioned otoliths have been used to age

many thousands of snapper whose otoliths were collected through market measuring in

1991, 1994 and then throughout the 2000s (Fowler and McGlennon 2011; Fowler et al.

2013). Growth conforms to the von Bertalanffy growth equation (McGlennon et al.

2000; SARDI unpublished data). The oldest snapper aged from SA was 36 years,

which is four years less than the oldest snapper captured in other Australian waters

(Norriss and Crisafulli 2010).

The many market-sampled snapper have been used to develop annual age structures

for the regional populations (Fowler et al. 2013). Such age structures display clear

strong and weak year classes that reflect the consequences of significant inter-annual

variation in recruitment of 0+ fish (McGlennon and Jones 1997; Fowler 2000, 2002;

Fowler et al. 2003, 2005, 2007, 2010; Fowler and McGlennon 2011). These age

structures indicate that the 1991 year class was particularly strong and largely

responsible for the recovery of the snapper fishery throughout all regions of the State,

following the poor catches of the 1980s. Further strong year classes recruited in 1997

Page 17: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

8

and 1999 and contributed significantly to all regional fisheries after 2000 (Fowler 2002;

Fowler et al. 2003, 2005, 2007, 2010, 2013).

1.3.3 Reproductive Biology

Snapper are multiple batch spawners that have indeterminate fecundity and

asynchronous oocyte development (Saunders 2009; Saunders et al. 2012). The

individual fish spawn over consecutive days, consistent with findings for Japan and New

Zealand (Matsuyama et al. 1988; Scott et al. 1993). Our best understanding for the

timing of reproductive activity is for NSG (Saunders 2009; Saunders et al. 2012). In this

region gonad development commences in October, spawning starts in late November,

peaks throughout December and then declines in January and is finished by early

February. The timing of gonad development and spawning appears to vary, to some

extent, amongst geographic regions, with reproductive activity persisting for longer in

the southern gulfs (SARDI unpublished data). Estimates of batch fecundity in NSG

ranged from 12,750 to >1,000,000 and up to 551,000 oocytes in SSG, and varied with

fish size, weight and ovary size. There is no evidence in SA that older snapper undergo

senescence of reproductive activity and output (SARDI unpublished data).

1.3.4 Early Life-History

Snapper eggs are approximately 0.85 – 1.0 mm in diameter, with unsegmented yolk

and a narrow perivitelline space (Crossland 1976). They are pelagic and hatch after

approximately 36 hours at 21°C, releasing larvae of 2 mm length, which remain pelagic

during development. The larvae take approximately 20 - 30 days to reach 8-12 mm

standard length (SL) when they are capable of undergoing settlement and becoming

demersal juveniles (Fukuhara 1985; Tanaka 1985; Fowler and Jennings 2003).

The distribution and abundance of 0+ snapper throughout NSG, were documented by

an annual recruitment survey from 2000 to 2010 (Fowler and Jennings 2003; Fowler et

al. 2005, 2007, 2010). The new recruits were distributed consistently amongst years,

being clumped both within and between adjacent stations that were dominated by bare,

flat, muddy substratum. This suggests that appropriate habitat for settlement is actively

selected by the 0+ fish. In both New Zealand and Japan, the 0+ fish are also strongly

associated with fine sediments typical of areas with low current regimes. Recruitment is

highly variable and until recently was thought to relate to survivorship of the larvae as

influenced by sea surface temperature (Fowler and Jennings 2003; Fowler et al. 2007).

However, further research has indicated that the controlling factors are more complex,

and relate to inter-annual variation in plankton trophic dynamics (Zeldis et al. 2005,

Page 18: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

9

Murphy et al. 2012, 2013). From the estimated spawn dates of the 0+ recruits, as

determined from the microstructure of their otoliths, it is apparent that successful

recruitment resulted from specific periods through the reproductive season, whose

duration and timing varied between years (Fowler et al. 2007).

1.4 Research program

SARDI’s research program for snapper involves core-funded monitoring that is

augmented by externally-funded projects to address specific issues. The core-funded

monitoring provides information on the demographic processes and population

dynamics of the regional populations. This is based on market-sampling of commercial

catches that are primarily accessed at Adelaide’s SAFCOL wholesale fish market.

Commercial catches are sampled from the various regions from which individual fish

are measured to develop size structures. The otoliths from some fish are removed to

determine age. The estimates of size-at-age are informative about growth and used to

generate age structures (Fowler and McGlennon 2011), which indicate annual

recruitment patterns (Fowler et al. 2013).

The core-funded monitoring work has also been augmented by specific projects to

elucidate particular aspects of snapper biology. These have included projects on adult

movement and stock structure (McGlennon 2004; Fowler et al. 2005b; Fowler 2016),

demographics including small-scale movement (Fowler and McGlennon 2011; Fowler

2016), reproductive biology, early life history and recruitment variability (Fowler and

Jennings 2003; Saunders et al. 2012), as well as fishery model development

(McGarvey and Feenstra 2004). A current FRDC-funded project is aimed at assessing

the applicability of the Daily Egg Production Method for estimating spawning biomass

for snapper populations in SA (SARDI unpublished data).

1.5 Information sources for stock assessment

1.5.1 Commercial catch and effort data

Since July 1983, all commercial fishers operating in SA’s MSF have been required to

submit a monthly catch return that details their fishing activity for the preceding month.

The required information includes catch weights by species, targeted fishing effort by

gear type, and the Marine Fishing Areas (MFAs) in which the fishing was done. Since

July 2003, it has been required that the month’s fishing activity be reported on a daily

basis. The returns are submitted to SARDI Aquatic Sciences where the data are

entered and stored in the Marine Scalefish Fisheries Information System. These catch

Page 19: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

10

and effort data constitute the most fundamental dataset available as indicators of

fishery status. The data available for this current assessment were from July 1983 to

December 2015, i.e. a 32.5 year dataset.

1.5.2 Recreational catch and effort data

Recreational fishery catches must be taken into consideration when determining fish

stock status and for resolving resource allocation issues. There have been four State-

wide recreational fishing surveys undertaken in SA to quantify recreational catch and

effort data. The first was a creel survey that was done through the period of 1994-96

(McGlennon and Kinloch 1997). Subsequently, there have been three telephone/ diary

surveys that were done in 2000/01 (Henry and Lyle 2003; Jones and Doonan 2005),

2007/08 (Jones 2009) and 2013/14 (Giri and Hall 2015). The four surveys provide

regional, species-specific estimates of recreational catch and targeted fishing effort.

1.5.3 Demographic data

Understanding the demographic processes that operate for the regional populations of

snapper has been facilitated by obtaining estimates of size-at-age. From these, size

and age structures have been developed and interpreted to provide estimates of

recruitment, growth and mortality rates. Since 2000, SARDI’s market sampling has

provided annual estimates of size-at-age of snapper. The emphasis on sampling

snapper has generally been lower during one of every three years, nevertheless at least

some snapper have been sampled in each year to 2015. Regional size and age

structures, growth curves, and recruitment histories were reported previously from

market sampling between 2000 and 2012 (Fowler and McGlennon 2011, Fowler et al.

2013).

1.5.4 ‘SnapEst’ stock assessment model

The computer model ‘SnapEst’ is a spatial, dynamic fishery model that was specifically

developed as a stock assessment tool for SA’s snapper fishery (McGarvey and

Feenstra 2004). The model provides time-series of biological fishery performance

indicators that are interpreted in terms of fishery status. Historically, the model has

recognised the three spatial regions of NSG, SSG and GSV that have traditionally

accounted for most of SA’s snapper catch. The snapper model employs the slice-

partition method to describe population numbers by both fish size and age (McGarvey

et al. 2007). It integrates all fishery and biological data available on population structure

and outputs time series of the parameters: fishable biomass; harvest fraction; egg

production; and recruitment. Stock status is determined from recent estimates of these

Page 20: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

11

output parameters that are assessed against biological performance indicators and

trigger reference points that are identified in the Management Plan for the South

Australian Commercial Marine Scalefish Fishery (PIRSA 2013).

1.6 Harvest strategy

1.6.1 Management plan

SA’s snapper fishery has experienced considerable change since 2007, particularly with

respect to the spatial structure of the fishery. Catches and catch rates in SG (the

traditional snapper region) have declined considerably, whilst those in NGSV and the

SE have increased to unprecedented levels. These changes resulted in several levels

of action: numerous management changes were implemented to limit commercial

catches; whilst several FRDC-funded research projects were undertaken to identify the

demographic processes responsible for the observed spatial changes (FRDC 2012/020,

Fowler 2016), and to develop a fishery independent index of fishable biomass (FRDC

2014/019).

The current harvest strategy for snapper (PIRSA 2013) recognises the changes

between 2007 and 2013 and the resulting challenges for fishery management. As

such, it does not include explicit decision rules with respect to responses to fishery

status. Rather, the proposed approach is to maintain the sustainability of the regional

fisheries whilst two FRDC-funded research projects (2012/020, 2014/019) are

completed. It proposes that the review of the Management Plan (PIRSA 2013) and

harvest strategy that is scheduled for 2018 take into consideration the enhanced

understanding from these two projects. The aim is to develop a better harvest strategy

with explicit decision rules for management responses to fishery status, based on

greater understanding of the biology and fishery, in order to provide greater certainty for

sustainable management.

1.6.2 Performance indicators

Determination of stock status for SA’s snapper fishery is assisted through assessment

of fishery performance indicators that are compared against trigger reference points.

There are two sets of fishery performance indicators, ‘general’ and ‘biological’ (PIRSA

2013, Table 1.1). The former are based entirely on commercial fishery statistics, whilst

the latter are largely based on output from the computer fishery model ‘SnapEst’, but

also include population age structures. In the harvest strategy, the fishery performance

indicators are differentiated into primary and secondary. The former are considered key

Page 21: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

12

determinants of fishery performance whilst the latter provide supporting information in a

weight-of-evidence assessment of fishery performance.

There are two other fishery performance indicators that are specific to snapper and

based on commercial fishery statistics, i.e. annual estimates of the proportions of

handline and longline fishing trips for which catches reached 250 kg. These

parameters are henceforth called ‘Prop250kgTarHL’ and ‘Prop250kgTarLL’, for

handlines and longlines, respectively. Their use became necessary due to the

imposition of a daily limit on commercial catch imposed in 2012, which reduced the

value of CPUE as an indicator of stock status.

1.6.3 Allocation amongst sectors

The Fisheries Management Act 2007 states that each sector in the fishery is entitled to

an allocation of the resource. For MSF species, the shares allocated amongst the

different sectors are those to which they had access at the time that the Minister

requested the Fisheries Council prepare the Management Plan, based on the most

recent information available (PIRSA 2013). At that time, the most recent data were

from 2007/08, when a State-wide recreational fishing survey had been done (Jones

2009). For the comparisons, trigger limits are specified in the Commercial and

Recreational Management Plans (PIRSA 2013, 2016b). They provide for some

variability in the proportional contributions to total catch between years, allowing some

flexibility for sectors to exceed allocations without triggering a review. Assessing

whether the different sectors have remained within their allocations is based on three

trigger limits. Trigger 1 relates to the comparison between commercial fisheries and the

recreational sector, and therefore can only be done when new recreational data are

available. It asks whether a sector’s contribution to total catch in the assessment year

exceeded its allocation by the percentage specified in Table 1.2. Triggers 2 and 3

relate only to the commercial fisheries. Because new commercial fishery data are

available each year they can be considered during each stock assessment process.

Trigger 2 considers whether a commercial fishery’s contribution to total commercial

catch exceeded its allocation by the percentage nominated in Table 1.2 in three

consecutive years or in four of the previous five years. Trigger 3 considers whether a

commercial fishery’s contribution exceeded its allocation of the total commercial catch

by the amount nominated in Table 1.2.

Page 22: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

13

Table 1.1 Fishery performance indicators and trigger reference points used to assess fishery performance as specified in the Management Plan (PIRSA 2013). The type of indicator and whether a primary or secondary one is also indicated. G – general; B – biological; P – primary; S – secondary.

Performance Indicator

Type P or S Trigger Reference Point

Total catch G S 3rd lowest/3rd highest

Greatest interannual change (±)

Greatest 5-year trend (±)

Decrease over 5 consecutive years?

Targeted handline effort

G P 3rd lowest/3rd highest

Greatest interannual change (±)

Greatest 5-year trend (±)

Decrease over 5 consecutive years?

Targeted longline effort

G P 3rd lowest/3rd highest

Greatest interannual change (±)

Greatest 5-year trend (±)

Decrease over 5 consecutive years?

Targeted handline CPUE

G P 3rd lowest/3rd highest

Greatest interannual change (±)

Greatest 5-year trend (±)

Decrease over 5 consecutive years?

Targeted longline CPUE

G S 3rd lowest/3rd highest

Greatest interannual change (±)

Greatest 5-year trend (±)

Decrease over 5 consecutive years?

Yearly prop. handline trips >250 kg

P 3rd lowest/3rd highest

Greatest interannual change (±)

Greatest 5-year trend (±)

Decrease over 5 consecutive years?

Yearly prop. longline trips >250 kg

S 3rd lowest/3rd highest

Greatest interannual change (±)

Greatest 5-year trend (±)

Decrease over 5 consecutive years?

Fishable biomass B P 3-yr ave is +/- 10% of previous 3-yr ave

Harvest fraction B P above 32% (int. standard)

Egg production B S <20% of pristine pop

Recruitment B S 3-yr ave is +/- 10% of historical mean

3-yr ave is +/- 10% of previous 6-yr ave

Age composition B P prop >10 yrs <20% of fished pop

Table 1.2 Allocation percentages and trigger limits for SA’s snapper fishery. Table shows the percentage of total allocation amongst commercial and recreational sectors, and trigger reference point for each sector (Trigger 1). It also shows commercial allocation and trigger reference points for Triggers 2 and 3. Fishing sectors are; MSF = Marine Scalefish, SZRL = Southern Zone Rock Lobster, NZRL = Northern Zone Rock Lobster, LCF = Lakes and Coorong, REC = general recreational, CHTR = charter boat, ABT = aboriginal/ traditional.

MSF SZRL NZRL LCF REC CHTR ABT

Fishery Allocation (%)

79.0 1.45 0.55 0.03 8.0 10.0 1.0

Trigger 1 (%) 84.0 2.9 1.65 1.0

Commercial allocation

97.5 1.78 0.68 0.04 - - -

Trigger 2 (%) na 2.68 1.3 0.75

Trigger 3 (%) na 3.58 2.0 1.0

Page 23: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

14

1.7 Stock status classification

A national stock status classification system was been developed for assessing

Australian fish stocks (Flood et al. 2014). It considers whether fishing pressure is

adequately controlled to ensure that stock abundance is not reduced to a point where

the production of juveniles is significantly compromised (Table 1.3). The system

combines information on both the current stock size and level of exploitation into a

single classification for each stock against defined biological reference points. Each

stock is then classified as either: ‘sustainable’, ‘transitional-recovering’, ‘transitional-

depleting’, ‘overfished’, ‘environmentally limited’, or ‘undefined’. PIRSA has adopted

this classification system for assigning status for all South Australian fish stocks. An

earlier version of this schema was used in the previous stock assessment for snapper

(Fowler et al. 2013).

Table 1.3 Classification scheme used to ascribe fishery stock status. The description of each stock status and its potential implications for fishery management are also shown.

Stock status DescriptionPotential implications for

management of the stock

Sustainable

Stock for which biomass (or biomass proxy) is at a level sufficient to ensure

that, on average, future levels of recruitment are adequate (i.e. not

recruitment overfished) and for which fishing pressure is adequately

controlled to avoid the stock becoming recruitment overfished

Appropriate management is in place

↑ Transitional–recoveringRecovering stock—biomass is recruitment overfished, but management

measures are in place to promote stock recovery, and recovery is occurring

Appropriate management is in

place, and the stock biomass is

recovering

↓ Transitional–depleting

Deteriorating stock—biomass is not yet recruitment overfished, but fishing

pressure is too high and moving the stock in the direction of becoming

recruitment overfished

Management is needed to reduce

fishing pressure and ensure that the

biomass does not deplete to an

overfished state

Overfished

Spawning stock biomass has been reduced through catch, so that average

recruitment levels are significantly reduced (i.e. recruitment overfished).

Current management is not adequate to recover the stock, or adequate

management measures have been put in place but have not yet resulted in

measurable improvements

Management is needed to recover

this stock; if adequate management

measures are already in place,

more time may be required for them

to take effect

Environmentally limited

Spawning stock biomass has been reduced to the point where average

recruitment levels are significantly reduced, primarily as a result of

substantial environmental changes/impacts, or disease outbreaks (i.e. the

stock is not recruitment overfished). Fisheries management has responded

appropriately to the environmental change in productivity

Appropriate management is in place

Undefined Not enough information exists to determine stock statusData required to assess stock

status are needed

Page 24: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

15

2. METHODS

2.1 Commercial fishery statistics

The catch and effort data for snapper were extracted from the commercial Marine

Scalefish Fisheries Information System which includes data from the MSF, NZRLF and

SZRLF. Data on snapper catch and effort were also extracted from the Lakes and

Coorong Fishery Information system and combined with the former dataset. Then, the

data from calendar years of 1984 to 2015, i.e. a 32-year time series were aggregated at

both the State-wide and regional spatial scales to provide annual estimates of total and

targeted catch, targeted effort for the two primary gear types of handline and longline.

Furthermore, the annual estimates of targeted catch and effort for the two gear types

were used to calculate catch per unit (CPUE). The time series of CPUE for the two

gear types in association with those for catch and effort, were qualitatively interpreted

here in terms of the regional trends in fishable biomass. Some regional time-series for

CPUE demonstrated long-term increasing trends that were thought to not relate to

increasing biomass, but to increasing ‘effective’ effort due to technological change. A

statistical approach was used to remove these long-term trends, as a possible way of

allowing for a more equitable comparison of CPUE over time. Each regional time-

series of CPUE was detrended by transforming the raw CPUE estimates using the

allometric relationship developed by Thorpe (1975):

�̂�𝑖 = 𝑙𝑜𝑔10𝑌𝑖 − 𝑏(𝑙𝑜𝑔10𝑋𝑖 − 𝑙𝑜𝑔10�̿�), where

�̂�𝑖 = the adjusted estimate of CPUE for the ith year

𝑌𝑖 = the unadjusted measurement of CPUE for the ith year

𝑏 = within-group regression coefficient of 𝑙𝑜𝑔10𝑌 against 𝑙𝑜𝑔10𝑋

𝑋𝑖 = year i

�̿� = overall mean year across the time series.

The data considered at the State-wide scale were ‘total catch’ and ‘targeted effort’

(mandays). Untargeted effort is not considered as it is not likely to be meaningful with

respect to stock status. The analysis of fishery statistics was considered here for six

regional management units (Fig. 2.1), consistent with the spatial scale considered

previously (Fowler et al. 2013), and is based on targeted fishing statistics for catch,

effort and CPUE for both handlines and longlines.

2.2 Recreational fishery statistics

There have been four State-wide recreational fishery surveys undertaken in South

Australia that now provide regional and State-wide estimates of recreational catch and

Page 25: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

16

effort. The first was a creel survey done in 1994-96 that used the ‘bus-route’ method for

surveying boat ramps (McGlennon and Kinloch 1997). The three subsequent surveys

used the telephone/diary methodology and were done in 2000/01 (Henry and Lyle

2003), 2007/08 (Jones 2009) and 2013/14 (Giri and Hall 2015).

2.3 Market sampling

Since 2000, market sampling of commercial catches has been undertaken, augmenting

similar work done for NSG in 1991 and 1994. The recent market sampling was

concentrated at the SAFCOL fish market although catches were occasionally accessed

from elsewhere. Generally, at this market, a team of three researchers processed

catches one morning per week prior to the morning auction. Catches of snapper were

selected for processing to ensure a broad geographic coverage. A two-stage sampling

protocol was used to process catches, where fish from each chosen catch were

measured for CFL (mm) and then otoliths were collected from a sub-sample of fish.

The latter fish were measured for CFL, weighed, gutted, and their sex and stage of

reproductive development were determined. Later, in the laboratory, one otolith from

each fish was embedded in resin and sectioned using a diamond saw to produce a thin

transverse section. The section was mounted on a glass microscope slide and then

examined using low power microscopy to count the opaque zones. This count, the

relative size of the most recent increment, and the time of year of capture, were used to

estimate the fish’s age (McGlennon et al. 2000). For each region in each year, an

age/length key was developed, based on the sizes and ages of fish that were aged

during that year, regardless of season or gear type.

Annual estimates of size and age structures were developed for those spatial

management units for which sufficient data were available, using the methods of

McGlennon et al. (2000). Development of each size structure was based on all fish

measured from handline and longline catches from each region, but weighted according

to the sizes of catches of the two gear types. The proportional length data were then

converted to a biomass distribution using a region-specific, length – weight relationship.

An age/length key was generated for each region and year, based on the size and age

of all fish aged during that year from each region, regardless of season and gear type.

Then, the age/length key was applied to the length frequency distribution to generate an

annual, region-specific age structure.

Page 26: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

17

Fig. 2.1 Map of South Australia showing the six regional management units that are considered in the stock assessment processes. Inset shows the location of the geographic region along Australia’s southern coastline. NSG – Northern Spencer Gulf, SSG – Southern Spencer Gulf, NGSV – Northern Gulf St. Vincent, SGSV – Southern Gulf St. Vincent, SE – South East, WC – West Coast.

2.4 Stock assessment model – SnapEst

The fishery stock assessment model SnapEst was developed in an FRDC-funded

project (McGarvey and Feenstra 2004), as a dynamic, spatial age- and length-

structured model. It integrates data collected from October 1983 to September 2015 to

estimate biological fishery performance indicators. Five data sets were used as input

data to the SnapEst model: commercial catch and effort data; recreational catch and

effort data from the telephone and diary surveys done in 2000/01, 2007/08 and

2013/14; charter boat catch and effort data from charter boat logbooks collected since

September 2005/06; commercial length-frequency samples; and commercial catch-at-

age samples.

SnapEst runs on a half-yearly time step, fitting to data from each summer (Oct-Mar) and

winter (Apr-Sep) time step since October 1983. The snapper model employs the slice-

partition method to describe fish population numbers by both age and length (McGarvey

et al. 2007; McGarvey et al. 2011). The model considers the catches and associated

effort divided into six categories (handline, longline, hauling net, all other commercial

gears combined, charter boats, and other recreational). Target type is not

differentiated, and non-target catches of snapper were, for the most part, low. Full

details of the model equations, fishery dynamics and likelihood function, are given in

Appendix 6.3.

SE

SGSV

NGSV

NSG

SSG

WC

Page 27: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

18

Spatially, the model is restricted to the two gulfs. Previously, the model recognised only

three spatial regions: NSG, SSG and GSV. In this assessment, GSV is divided into

NGSV and SGSV based on the recent evidence of different demographic influences in

the two regions (Fowler 2016). These model regions are consistent with the spatial

management units for which commercial fishery data are also assessed (Section 2.1).

The model changes and additional prior analyses of growth (length-at-age) and weight-

length for these two new GSV regions are presented in Appendix 6.4. The WC and SE

regions are not included in the model because of low historical catches and low sample

sizes from market sampling. There is no movement assumed in the model. Thus, the

different model regions are considered to be independent of each other.

SnapEst integrates the five input data sets, to carry out maximum likelihood estimation

of four annual indicators of fishery performance: recruitment rates, fishable biomass,

harvest fraction, and egg production. Recruitment and egg production are computed

once per model year while annual indicators for biomass and harvest fraction are

computed from the two half-yearly model estimates. Yearly recruit numbers (i.e.

numbers reaching 2 years of age at the start of the model half year that commences in

October) are dated by year class from the summer when spawned. Annual fishable

biomass is the average of the two half-yearly estimates, which start in October and April

of each model year. Harvest fraction is the sum of model-predicted half-yearly catches

divided by the annual fishable biomass. The proportion of pristine egg production is the

estimated annual egg production (assuming a 50:50 sex ratio, and fecundity and

maturity ogives divided by a single equilibrium estimate of ‘pristine’, i.e. pre-fishery, egg

production obtained by running the model as a projection with fishing effort set equal to

zero for all years.

2.5 Assessment of fishery performance

2.5.1 Fishery Performance Indicators

The fishery performance indicators for snapper (Table 1.1) were assessed at the spatial

scale of regional spatial management unit (Fig. 2.1). The general fishery performance

indicators considered for each region were; total catch, targeted handline effort,

targeted longline effort, targeted handline CPUE and targeted longline CPUE. For each

region the time series of data from 1984 to 2015 for each indicator was calculated.

Then, the value for 2015, was compared using the trigger reference points calculated

Page 28: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

19

for the ‘reference period’, i.e. the historical data time series back to 1984 (Appendix 4,

PIRSA 2013). This comparison was done by addressing four questions:

was the value of the indicator in 2015 among either the top three or bottom three

values over the reference period of 1984 to 2015?

was the change in the indicator between 2014 and 2015, i.e. the two most

recent years, the greatest inter-annual increase or decrease over the reference

period?

was the slope of the linear trend over the last five years to 2015, the greatest

rate of increase or decrease over five-year periods throughout the reference

period?

and did the indicator decrease over the last five consecutive years?

A ‘results’ table was prepared that showed the outcomes of these comparisons for the

six regions, indicating whether or not the target reference points had been breached.

Furthermore, the proportions of handline and longline fishing trips for which catches

reached 250 kg were also considered fishery performance indicators, i.e.

‘Prop250kgTarHL’ and ‘Prop250kgTarLL’ . The calculation of these performance

indicators was based on daily catch data from the commercial sector. The requirement

to report such data on a daily rather than monthly basis only became mandatory in July

2003. As such, the calculation of these two performance indicators was restricted to

the calendar years of 2004 to 2015. Furthermore, the targeted catch data from

November to January in each summer were excluded from these calculations in order

to remove the influence of the seasonal closure on the data (PIRSA 2013). The

regional estimates of Prop250kgTarHL and Prop250kgTarLL for 2015 were compared

against those from 2004 to 2014, using the same trigger reference points that are used

for the general performance indicators.

There are five biological performance indicators: fishable biomass; egg production;

harvest fraction; recruitment; and age structure (Table 1.1; PIRSA 2013). The first four

are time-series of output parameters from ‘SnapEst’, whilst the age structures are catch

proportions by age from market sampling. The estimates of output parameters were

considered for the four model regions considered in SnapEst, i.e. NSG, SSG, NGSV

and SGSV. The trigger reference points are indicated in Table 1.1. For the recruitment

indicator, only model estimates up to and including the 2011 year class are included in

this assessment, as it takes several years for a year class to fully recruit to the fishery.

For the recruitment performance indicator, the most recent 3-year average covering

Page 29: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

20

2009-2011 was compared with the average over the previous 6 year classes, 2003-

2008, and with the historical mean taken over 1983-2008.

2.5.2 Comparison with allocations for fishery sectors

For this assessment, new recreational fishery data from the survey undertaken in

2013/14 (Giri and Hall 2015), allowed contributions of the four commercial fisheries and

the recreational sector to be assessed against their allocations and trigger limits

(Trigger 1 – Table 1.2). The reported catches and allocations for the commercial sector

from 2013/14 that were compared against the data from the recreational survey period

of December 2013 to November 2014. For the assessment of allocation amongst only

the commercial fisheries, Triggers 2 and 3 (Table 1.2) were used. The data considered

here were from the calendar year of 2015.

3. RESULTS

3.1 Commercial fishery

3.1.1 State-wide fishery statistics

Estimates of total State-wide commercial catch of snapper show cyclical variation, with

the cycles typically increasing and decreasing over several years (Fig. 3.1a). Since

2003, i.e. the year that produced the minimum catch at the start of the most recent

cycle, State-wide catch increased to a record level of 1,035 t in 2010, before declining

by more than 50% to 512 t in 2015. Handlines were historically the most significant

gear type, largely accounting for cyclical variation in total catch until 2008 (Fig. 3.1a).

Longline catch increased marginally between 2005 and 2008 before increasing

considerably until 2010, when it became the dominant gear type. Both longline and

handline catches declined between 2010 and 2013, but were relatively consistent

between 2013 and 2015. The commercial catch has historically been dominated by the

MSF which has generally contributed >95% of the reported catch (Fig. 3.1a). The

NZRLF has generally contributed <3% of the total, but has been as high as 7%. That

for the SZRLF has generally been <1% of total catch, but ranged from 2 to 4% between

2008 and 2013. The annual catch of the LCF has always been very low and never

exceeded 0.4% of total catch.

There was a long-term declining trend of targeted commercial fishing effort between the

mid-1980s and 2008 (Fig. 3.1b). This was followed by a period of elevated fishing effort

between 2009 and 2012 that was associated with the increase in longline effort.

Page 30: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

21

However, since 2011, longline effort has declined exponentially, complementing the on-

going, long-term declining trend for handline effort. As such, in 2015, targeted fishing

effort was the lowest since 1984. The MSF has generally accounted for >94% of

targeted effort. Estimates from the NZRLF and SZRLF have accounted for up to 6%

and that of the LCF has generally been <0.5% of targeted effort.

The number of fishers from across all four commercial sectors who reported taking

snapper, declined consistently from 401 in 1984 to 244 in 2000. It then stabilised for a

period before once again declining from 2010 onwards. Between 2010 and 2015, it fell

from 266 to 186 fishers. The number of commercial fishers who targeted snapper

varied similarly and fell to the lowest number of 150 in 2015.

Fig. 3.1 State-wide annual commercial fishery statistics by calendar year from 1984 to 2015. a. catches of snapper by commercial fishery and gear type. b. estimates of targeted fishing effort by fishery and gear type. c. numbers of license holders who reported taking and targeting snapper in each year.

Page 31: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

22

3.1.2 Regional Analysis of Catch and Effort Statistics (1984 – 2012)

The division of the State-wide, annual estimates of total catch into their regional

components produced temporal trends that differed amongst regions and indicated a

significant change in the spatial structure of the fishery over time (Fig. 3.2). NSG

supported the highest regional catches from 1984 to 2004, which subsequently declined

and fell to their lowest levels between 2012 and 2015. SSG was the dominant regional

contributor to annual State-wide catch between 2005 and 2009, after which regional

catches declined rapidly. Catches in NGSV were very low until around 2004, increased

gradually for a few years before this accelerated between 2007 and 2010. This region

has remained the dominant contributor to total catch up to 2015. The catches in the SE

also increased dramatically between 2007 and 2010 before declining consistently to a

relatively low level in 2015. The annual catches from SGSV and the WC have generally

been relatively minor. SGSV experienced a marginal increase from 2009 to 2011. The

highest catches for the WC were taken in 2008 and 2009 and have since fallen

considerably.

Fig. 3.2 Relative estimates of total annual commercial catch of snapper by region, indicated by sizes of circles.

Page 32: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

23

Northern Spencer Gulf

Total catch from NSG has varied cyclically with peaks in 1990, 1999 and 2010, and low

catches recorded in 1987, 1994 and 2005 (Fig. 3.3a). The highest catch of 360 t was

recorded in 1999, after which there has been a general decline, particularly in 2012.

Low catches persisted until 2015 when it was 47 t. Targeted handline catch was

highest between 1998 and 2002 with the highest of 319 t taken in 1999, which fell to

only 40 t in 2015 (Fig. 3.3b). Targeted handline fishing effort has also decreased since

2002, with particular drops in 2005 and 2012 (Fig. 3.3c). Between 2002 and 2015, it fell

from 3,405 to 494 fisherdays, a decline of 85.5%. From 1984 to 2011, targeted CPUE

showed a long-term increasing trend, although varying cyclically in association with

targeted catch (Fig. 3.3d). Then in 2012, it declined by 51.4% from 140.6 to 68.4

kg.fisherday-1. It declined further in 2013, before increasing marginally in 2014 and then

to 79.9 kg.fisherday-1 in 2015. The detrended estimates of CPUE varied cyclically

through the 32-year time-series, with the value for 2013 the lowest ever, whilst those for

2014 and 2015 remained amongst the lowest. The number of license holders who

targeted snapper has also shown a long-term decline from the highest number of 88 in

1992 to 28 in 2015, with the lowest of 24 in 2014 (Fig. 3.3e). The estimates of

Prop250kgTarHL were variable, and relatively high in 2014 and 2015 (Fig. 3.3f).

Targeted longline catch peaked at 84 t in 1997 (Fig. 3.3g). It decreased regularly after

that and in 2015 was at only three tonnes, the lowest yet recorded. Longline fishing

effort declined from 2,353 fisherdays in 1997 to only 49 fisherdays in 2015, i.e. a

reduction of 97.9% in 18 years (Fig. 3.3h). Targeted longline CPUE was relatively

consistent until 2002 (Fig. 3.3i). Subsequently, it was highly variable, varying between

87.3 kg.fisherday-1 in 2007 and 24.3 kg.fisherday-1 in 2013. It increased to 50.5

kg.fisherday-1 in 2015, although the catch and effort levels from which the values were

calculated in 2012 to 2015 were very low. The detrended estimates of CPUE in 2013

and 2014 were the lowest ever, but did increase in 2015. The number of longline

fishers who targeted snapper declined consistently from 1993, and in 2015 was only 5

fishers (Fig. 3.3j). Prop250kgTarLL was low from 2010 to 2015, but increased in 2015,

although the number of catches was very low (Fig. 3.3k).

Page 33: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

24

Fig. 3.3 Northern Spencer Gulf. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1).

Page 34: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

25

Southern Spencer Gulf

After low catches taken in SSG during the mid-1990s, total catch demonstrated two

modes during the 2000s, the first peaked in 2001 at 294 t, whilst the second had a

maximum of 402 t in 2007 (Fig. 3.4a). Total catch has subsequently declined regularly,

falling to the lowest ever reported level in 2014 and 2015. Targeted handline catch has

been highly variable since the low catch of 1994 (Fig. 3.4b). It peaked at 251 t in 2001

and again at 236 t in 2007. Since then it fell to 14 t in 2012, and further to only 6 t in

2015, the lowest recorded level. Targeted handline fishing effort has also varied

cyclically, increasing from 559 fisherdays in 1994 to a peak of 1,631 fisherdays in 2001

(Fig. 3.4c). It has subsequently declined to the lowest level of only 106 fisherdays in

2015. Targeted CPUE peaked at 154 kg.fisherday-1 in 2001, and then at the record

level of 182.6 kg.fisherday-1 in 2007 (Fig. 3.4d). However, from then until 2015, it

declined by 68.8% to 56.9 kg.fisherday-1. The detrended estimates of CPUE in 2013,

2014 and 2015 were comparable to the low values in 1994 and 1995. The numbers of

fishers targeting snapper have been highly variable (Fig. 3.4e). They were particularly

low through the mid-1990s, increased to their highest number of 86 in 2002, but then

decreased consistently to the lowest number of 21 in 2015. Prop250kgTarHL increased

in 2014 and again in 2015 (Fig. 3.4f).

Targeted longline catch in SSG was very low until 2004, then increased rapidly in 2005

and peaked at 127.0 t in 2006 (Fig. 3.4g). It then declined substantially to only 10.3 t in

2015. The increase in catch between 2005 and 2009 was associated with a substantial

increase in targeted fishing effort from 274 fisherdays in 2004 to 985 fisherdays in 2007,

i.e. an increase of 259.5% (Fig. 3.4h). Nevertheless, since then, targeted effort has

declined and in 2014 and 2015 was at the lowest level since the late 1990s. Targeted

CPUE increased gradually through the late 1990s and early 2000s to around 70

kg.fisherday-1 (Fig. 3.4i). In 2005, it increased sharply to >120 kg.fisherday-1, and

remained at this high level until 2008. It has subsequently fallen considerably to the low

value of 52.0 kg.fisherday-1 in 2014. The detrended estimates of longline CPUE in 2014

and 2015 were the lowest ever, even compared to the low values of the mid-1990s (Fig.

3.4i). The number of longline fishers that operated in SSG was highest at 31 in 2007,

when record catches and catch rates were taken and effort levels were highest (Fig.

3.4j). They have since decreased consistently to 16 fishers in 2015. Prop250kgTarLL

has declined since 2008 and was low in both 2014 and 2015 (Fig. 3.4k).

Page 35: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

26

Fig. 3.4 Southern Spencer Gulf. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1).

Page 36: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

27

Northern Gulf St. Vincent

Historically, NGSV produced only very low catches from the 1980s to early 2000s (Fig.

3.5a). However, from 2007 to 2010, total catch increased exponentially, culminating in

the record catch of 417 t in 2010. This has subsequently declined marginally to 380 t in

2015. Targeted handline catch in this region was generally low at <15 t.yr-1 up to 2006

and peaked at 46.5 t.yr-1 in 2010 (Fig. 3.5b). It has subsequently declined consistently

to 7.5 t.yr-1 in 2015. The higher catches to 2010 were associated with an increase in

targeted handline effort from 2004 to 2010 (Fig. 3.5c). This has subsequently declined

to 147 fisherdays in 2015. Targeted handline CPUE was relatively stable until 1996,

after which it was variable, but nevertheless showed a long-term increase to 2012 (Fig.

3.5d). Since then it has declined by 49.0% to 50.9 kg.fisherday-1 in 2015. Detrended

CPUE declined steeply from 2012 to its lowest level in 2015 (Fig. 3.5d). The number of

handline fishers that operated in NGSV increased in 2004 and remained high until

2012, before declining to less than half in 2015 (Fig. 3.5e). Estimates of

Prop250kgTarHL increased from 2004 to 2013 before declining in 2014 and 2015 (Fig.

3.5f).

The longline fishery in NGSV has largely accounted for the recent rapid increase in total

catch. Between 2008 and 2012, targeted longline catch increased by 873.5% from 37.0

to 360.2 t, and continued to increase to 370.6 t in 2015 (Fig. 3.5g). This increase was

associated with a 542.1% increase in longline fishing effort from 390 to 2,504

fisherdays.yr-1 in 2015 (Fig. 3.5h). Longline CPUE demonstrated a long-term increase

primarily between 2000 and 2010, when it peaked at 161 kg.fisherday-1 (Fig. 3.5i). It

has marginally declined to 148 kg.fisherday-1 in 2015. This recent decline was

accentuated in the detrended estimates of CPUE, which suggest a considerable recent

fall in catch rate (Fig. 3.5i). The number of fishers who used longlines to target snapper

in NGSV increased from 11 in 2007 to 50 in 2011 and 2012, but has declined to 36 in

2015 (Fig. 3.5j). Prop250kgTarLL increased considerably from 2004 to 2010 and was

highest in 2014 and 2015 (Fig. 3.5k).

Page 37: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

28

Fig. 3.5 Northern Gulf St. Vincent. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1).

Page 38: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

29

Southern Gulf St. Vincent

This regional fishery in SGSV is characterised by relatively low catches (Figs. 3.2,

3.6a). The highest catches of around 80 t were taken through the 1980s, but have

subsequently ranged from 20 – 40 t. Such catches are approximately one tenth of

those of the more productive regions. In the period of 2009 to 2011, relatively high

annual catches of around 38 t were taken, but they have subsequently declined to 27 t

in 2015. The targeted handline fishery in SGSV produced its highest catches in the

1980s and early 1990s, before dropping to a low of only 4.7 t in 1995 (Fig. 3.6b). Since

then there has been some cyclical variation but nevertheless it has remained low at <20

t.yr-1, particularly between 2008 and 2015. Targeted handline effort also declined

between 1984 and 1995, and has subsequently remained low, particularly 2008 and

2015 (Fig. 3.6c). Targeted CPUE was variable but increased consistently between

1995 and 2007, reaching a peak of 48.2 kg.fisherday-1 (Fig. 3.6d). It declined

significantly to 19.1 kg.fisherday-1 in 2012 before increasing again to 33.4 kg.fisherday-1

in 2015. Detrended CPUE showed two periods of low CPUE, i.e. through the mid to

late-1990s and from 2010 to 2012 (Fig. 3.6d). It was considerably higher in 2014 and

2015. The number of handline fishers who targeted snapper declined from 70 in 1985

to 17 in 2000, and has remained low until 2015 (Fig. 3.6e). Prop250kgTarHL has

always been low, and since 2012 was at zero (Fig. 3.6f).

Targeted longline catch remained at <10 t.yr-1 until 2009 when it increased to 18.4 t and

then again to 27.5 and 28.6 t in the following two years (Fig. 3.6g). From 2012 to 2015,

it has been lower at 13 – 17 t.yr-1. Effort was relatively stable at <250 fisherdays.yr-1

between 1997 and 2004, but increased to 549 fisherdays in 2011 (Fig. 3.6h),

corresponding to the period of expansion of longline fishing in NGSV. Nevertheless, it

has subsequently declined to 344 fisherdays in 2015. Targeted longline CPUE

demonstrated a general increasing trend between 1996 and 2010 from 15.2

kg.fisherday-1 to a record level of 68.5 kg.fisherday-1 (Fig. 3.6i). It subsequently

declined to 31.5 kg.fisherday-1 in 2012 before increasing to 50.6 kg.fisherday-1 in 2015.

Detrended CPUE shows the recent estimates to be at a moderate level. The number of

longline fishers was as high as 25 between 2010 and 2012, but has subsequently

declined to 19 in 2015 (Fig. 3.6j). Prop250kgTarLL has been variable with peaks on

2010 and 2014, and remained relatively high in 2015 (Fig. 3.6k).

Page 39: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

30

Fig. 3.6 Southern Gulf St. Vincent. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1).

Page 40: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

31

South East

Historically, the SE region produced only marginal catches of snapper (Fig. 3.7a).

However from 2006 to 2010 there was an exponential increase that culminated in the

record catch of 271 t in 2010. Since then, it has declined regularly back to 16 t in 2015.

Targeted handline catch in the SE has always been low (Fig. 3.7b). There was a minor

increase from 2006 to 2009, which has subsequently declined to <1 t in 2015. Such

catches reflect low but variable fishing effort, which declined from 287 fisherdays in

2007 to 28 in 2015 (Fig. 3.7c). CPUE has generally been less than 20 kg.fisherday-1,

but increased from 2003 and was highest from 2006 to 2009. By 2012, it had declined

to a minimum level (Fig. 3.7d). The detrended levels of CPUE were highest between

2006 and 2009 (Fig. 3.7d). The number of fishers who targeted snapper was variable

up to 2009, although generally <10 before declining to only two fishers in 2015 (Fig.

3.7e). Prop250kgTarHL was highest from 2006 to 2009, and has subsequently been at

very low levels (Fig. 3.7f).

Targeted longline catches were always less than several tonnes.yr-1 up to 2007 (Fig.

3.7g). From then, there was a rapid increase to the maximum level of 239 t in 2010,

which has subsequently declined to 15 t in 2015. There was a considerable increase in

targeted longline effort, which peaked in 2010 at 2,614 fisherdays, but which has

subsequently declined to 291 fisherdays (Fig. 3.7h). Targeted CPUE also increased

dramatically from 2007 to 2010, peaking at 91.5 kg.fisherday-1 before declining to 51.8

kg.fisherday-1 in 2015 (Fig. 3.7i). The detrended estimates of longline CPUE also

indicate a considerable decrease between 2010 and 2015 (Fig. 3.7i). The number of

longline fishers who targeted snapper increased dramatically from 2005 and peaked at

25 in 2010 before declining gradually to 11 fishers in 2015 (Fig. 3.7j). Prop250kgTarLL

was relatively high from 2007 to 2014, before declining in 2015 (Fig. 3.7k).

Page 41: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

32

Fig. 3.7 South East. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1).

Page 42: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

33

West Coast

The WC region has historically produced relatively low catches of snapper,

nevertheless with some modes in production (Fig. 3.8a). The first peaked in the 1980s

and then declined to low levels in the mid-1990s. The latter peaked in 2008 at 48 t.

Since then, total catch has declined to 18 t in 2015. Targeted handline catch from the

WC has generally been <12 t.yr-1 (Fig. 3.8b). It peaked in 2008 at 18.5 t.yr-1 and has

subsequently fallen to only 3.0 t.yr-1 in 2015. Targeted handline effort has also been

quite variable, but since 2005 has declined from 338 to 90 fisherdays (Fig. 3.8c).

Targeted CPUE increased from 1995, peaking in 2008 and 2009 (Fig. 3.8d). It has

subsequently fallen to 32.9 kg.fisherday-1 in 2015. The recent estimates of detrended

CPUE are historically low, with only those from the mid-1990s being lower. The

number of fishers who targeted snapper with handlines has shown two longterm cycles

(Fig. 3.8e). There was a high period during the 1980s, when the number reached 32,

which declined back to below 20 during the 1990s. They increased again in the late

1990s and early 2000s, peaking at 34 in 2001. Since then there has been a gradual

long-term decline back to below 20 fishers. Prop250kgTarHL was generally low,

particularly from 2010 to 2015 (Fig. 3.8f).

Targeted longline catch was relatively low until the early 2000s when it increased to

peak at 25.7 t.yr-1 in 2007 (Fig. 3.8g). Targeted effort displayed similar variation,

increasing to a maximum of 605 fisherdays in 2009, before declining to 340 fisherdays

in 2015 (Fig. 3.8h). Targeted CPUE increased after 1998 reaching a maximum of 93.5

kg.fisherday-1 in 2003 (Fig. 3.8i). It has decreased systematically since then to 31.7

kg.fisherday-1 in 2015. Detrended CPUE showed a similar decline between 2003 and

2015 (Fig. 3.8i). The number of longline fishers varied considerably between 1984 and

2015 (Fig. 3.8j). However, the number was highest between 2006 and 2015, peaking in

2008 at 29 fishers, and has subsequently declined to 23 fishers in 2015.

Prop250kgTarLL has always been low for this region (Fig. 3.8k).

Page 43: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

34

Fig. 3.8 West Coast. a. Time series of total catch. Left hand graphs – time series for targeted handline sector. b. catch. c. effort. d. raw (black line) and detrended (grey line) CPUE. e. number of license holders. f. Prop250kgTarHL (line graph) and number of catches (bar chart). Right hand graphs – time series for targeted longline sector. g. catch. h. effort. i. raw (black line) and detrended (grey line) CPUE. j. number of license holders. k. Prop250kgTarLL (line graph) and number of catches (bar chart). Crosses indicate confidential data, i.e. < 5 fishers. Third highest (green, blue lines) and third lowest (red, brown lines) values are shown for fishery performance indicators (Table 1.1).

Page 44: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

35

3.2 Recreational fishery

The results from the three recreational telephone /diary surveys that were undertaken in

2000/01, 2007/08 and 2013/14 were compared (Henry and Lyle 2003, Jones 2009, Giri

and Hall 2015). The estimated number of snapper harvested by the recreational sector

in 2013/14 was 207,809 individuals, which was the highest amongst the three surveys

(Fig. 3.14a). The numbers were dominated by catches from Gulf St. Vincent (Fig.

3.14b), which differs from the 2000/01 survey, when catches from SG gave the highest

numbers. The average estimated weight per individual fish in 2013/14 was 1.6 kg,

which is considerably lower than for the previous surveys (Fowler et al. 2010). In

2013/14, the total estimated weight of snapper captured by the recreational sector was

332.5 t, which was intermediate between those of the two earlier surveys (Fig. 3.14c).

This included charter boat catch based on fishers surveys, but not from charter boat

logbooks. The distribution of catch amongst regions was dominated by that from GSV,

with relatively low amounts captured from NSG, SSG, SE and WC (Fig. 3.14d). The

proportion of total catch accounted for by the recreational sector varied between the

three surveys. In 2013/14, 38.4% of total catch was attributable to the recreational

sector, which is intermediate relative to the estimates of 42.6% in 2001/01 and only

19.6% in 2007/08 (Fig. 3.14e). There was considerable regional variation evident for

each survey in the contribution to total catch by the recreational sector (Fig. 3.14f).

Fig. 3.9 Recreational fishery statistics from three State-wide telephone diary surveys.

Page 45: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

36

3.3 Stock assessment model - SnapEst

The yearly estimates of SnapEst biological performance indicators for the four modelled

regions are shown in Figures 3.10 to 3.13. With 95% confidence interval error bars,

model outputs are shown for fishable biomass, recruitment numbers, yearly harvest

fraction, and egg production, the latter expressed as a percentage of an unfished

(‘pristine’) stock.

Declining levels of harvest fraction in the two Spencer Gulf regions (Figs. 3.10, 3.11)

principally reflected reductions in targeted fishing effort. Relatively high estimates of

biomass in NSG and SSG reflect these lower exploitation rates, along with model-

inferred better recruitment in NSG in the last three year classes shown of 2009-2011.

Model estimates of much higher biomass and higher exploitation levels in NGSV reflect

similar conclusions from catch and effort data in that region of all-time high snapper

catches and catch rates. The majority of South Australian snapper fishing is now

focused in this high-catch region. Biomass estimates in SGSV may be optimistic.

Page 46: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

37

Fig. 3.10 Time series of the four annual biological performance indicators from the SnapEst fishery assessment model for Northern Spencer Gulf. Green lines show averages over the last three years, compared with blue lines giving averages over the three preceding years for biomass and six preceding years for recruitment, as in the trigger reference points for these indicators in Table 3.2.

Page 47: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

38

Fig. 3.11 Time series of the four annual snapper biological performance indicators from the SnapEst model for Southern Spencer Gulf.

Page 48: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

39

Fig. 3.12 Time series of the four annual snapper biological performance indicators from the SnapEst model for Northern Gulf St. Vincent.

Page 49: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

40

Fig. 3.13 Time series of the four annual snapper biological performance indicators from the SnapEst model for Southern Gulf St. Vincent.

Page 50: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

41

3.4 Year class contributions to fishery catches

The relative contributions to fishery catches of year classes from different time periods

were considered for the six regional populations, based on the age structures presented

in the Appendix (Figs. 6.2, 6.4, 6.6, 6.8, 6.10). For NSG, up to 2011, year classes from

the 1990s contributed >40% of the numbers of fish taken in the fishery, which fell to

20% in 2015. For SSG, >50% of fish captured in each year through the 2000s came

from year classes from the 1990s. In comparison, for NGSV, by 2011 <20% of the fish

captured were from year classes from the 1990s. As such, most of the large catches

taken from 2010 to 2015 had recruited as 0+ fish throughout the 2000s. For both

SGSV and the SE from 2010 onwards, the catches involved no fish from any year

classes prior to the 2000.

Fig. 3.14 Estimates of annual total catches for each region and percentage of population that recruited prior to 2000, determined from annual age structures.

Page 51: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

42

3.5 Fishery performance

3.5.1 Assessment against allocations

The availability of new recreational fishery data from the survey undertaken in 2013/14

(Giri and Hall 2015), made it possible to assess the contributions of the four commercial

fisheries and the recreational sector against the allocations and trigger limits that are

specified in Table 1.2. The four commercial fisheries each took less than their

allocations (Table 3.1). Alternatively, the percentage of total catch taken by the

recreational sector exceeded its allocation and trigger limit.

The assessment of the commercial catches in 2015 against Triggers 2 and 3 indicated

that no allocations were exceeded.

Table 3.1 Comparisons of percentages of total catch of snapper taken by the different fishery sectors with their allocations and trigger limits specified in the Management Plan (PIRSA 2013). Trigger 1 relates to commercial and recreational sectors. Triggers 2 & 3 relate only to commercial fishery data. MSF – Marine Scalefish, SZRL – Southern Zone Rock Lobster, NZRL – Northern Zone Rock Lobster, LCF – Lakes and Coorong. Green colour – allocation not exceeded, red colour – allocation trigger activated.

MSF SZRL NZRL LCF REC CHTR ABT

Fishery Allocation (%)

79.0 1.45 0.55 0.03 8.0 10.0 1.0

Trigger 1 (%) 84.0 2.9 1.65 1.0 27.0

% total 2014 61.0 1.06 0.08 0.15 37.7

Commercial allocation

97.5 1.78 0.68 0.04 - - -

Trigger 2 (%) na 2.68 1.3 0.75

Trigger 3 (%) na 3.58 2.0 1.0

% total 2011 96.6 2.3 0.7 0.3

% total 2012 97.3 2.3 0.3 0.1

% total 2013 96.5 3.1 0.4 0.1

% total 2014 98.9 0.7 0.1 0.3

% total 2015 99.4 0.5 0.2 0

3.5.2 General Performance Indicators

The general fishery performance indicators, and also ‘Prop250kgTarHL’ and

‘Prop250kgTarLL’, were assessed at the regional spatial scale. There were 18

breaches of trigger reference points across the six regions. For NSG, three triggers

were breached that related to low estimates of total catch and targeted handline and

longline effort (Table 3.2, Fig. 3.3). For SSG in 2015, the lowest handline effort was

recorded (Fig. 3.4), and also the highest estimate of ‘Prop250kgTarHL’ (Table 3.2, Fig.

Page 52: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

43

3.4). The latter related to a very low number of catches being taken from SSG in this

year.

For NGSV in 2015, there were five trigger reference points that were breached (Table

3.2). The second highest longline effort was recorded (Fig. 3.5). Also, there were

contrasting results for ‘Prop250kgTarHL’ and ‘Prop250kgTarLL’. The former was low

having decreased considerably compared to 2014 (Fig. 3.5), whilst, the latter was the

highest ever recorded. For SGSV in 2015, there were several breaches of trigger

reference points (Table 3.2). The lowest handline effort was recorded (Fig. 3.6). There

were no fishing trips that produced handline catches that exceeded 250 kg.

Alternatively, the longline sector produced its third highest value of ‘Prop250kgTarLL’

(Fig. 3.6).

For the SE, there were two breaches of trigger reference points for fishery data from

2015 (Table 3.2). Both total catch and targeted longline effort decreased over five

consecutive years between 2010 and 2015 (Fig. 3.7). For the WC, there were also two

breaches of trigger reference points in 2015 (Table 3.2). Targeted handline effort was

the third lowest, whilst ‘Prop250kgTarHL’ was zero (Fig. 3.8).

3.5.3 Biological Performance Indicators (BPIs)

The time series of plotted model-estimated BPIs (Figs. 3.10 to 3.13) show generally

stable or rising trends in abundance. Accordingly, model-based BPIs yielded

predominantly positive directions when they triggered. Quite high recruitment variation

of South Australian snapper stocks gives rise to wide variation in the performance

indicators, which was evident for the past three years of this assessment. For this

reason, averages over the past three years were compared with averages over the

preceding 3 years (biomass) or the preceding 6 years and the historical average

(recruitment). Model biomass levels may be biased on the high side in Spencer Gulf

and SGSV, meaning exploitation rates are higher than shown.

Because of the high variability in snapper recruitment, recruitment indicators triggered

in all four model regions, being more than 10% higher or lower than the previous 3

years, or the historic averages.

Page 53: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

44

Table 3.2 Fishery performance indicators and trigger reference points used to assess fishery performance as identified in the Management Plan (PIRSA 2013). The type of indicator and whether a primary or secondary one is also indicated. G – general; B – biological; P – primary; S – secondary.

Performance Indicator Type P or S Trigger Reference Point NSG SSG NGSV SGSV SE WC Total catch G S 3rd lowest/3rd highest 3rd lowest 2nd lowest n n n n

Greatest interannual change (±) n n n n n n

Greatest 5-year trend (±) n n n n n n

Decrease over 5 consecutive years? n n n n Y n

Targeted handline effort G P 3rd lowest/3rd highest Lowest Lowest n Lowest n 3rd lowest

Greatest interannual change (±) n n n n n n

Greatest 5-year trend (±) n n n n n n

Decrease over 5 consecutive years? n n n n n n

Targeted longline effort G P 3rd lowest/3rd highest Lowest n 2nd high n n n

Greatest interannual change (±) n n n n n n

Greatest 5-year trend (±) n n n n n n

Decrease over 5 consecutive years? n n n n Y n

Targeted handline CPUE G P 3rd lowest/3rd highest n n n n n n

Greatest interannual change (±) n n n n n n

Greatest 5-year trend (±) n n n n n n

Decrease over 5 consecutive years? n n n n n n

Targeted longline CPUE G S 3rd lowest/3rd highest n n n n n n

Greatest interannual change (±) n n n n n n

Greatest 5-year trend (±) n n n n n n

Decrease over 5 consecutive years? n n n n n n

Yearly prop. handline trips >250 kg

P 3rd lowest/3rd highest n Highest 3rd lowest Y-zero n Y-zero

Greatest interannual change (±) n n Highest ↓ n n n

Greatest 5-year trend (±) n n Highest ↓ n n n

Decrease over 5 consecutive years? n n n n n n

Yearly prop. longline trips >250 kg

S 3rd lowest/3rd highest n n Highest 3rd high n n

Greatest interannual change (±) n n n n n n

Greatest 5-year trend (±) n n n n n n

Decrease over 5 consecutive years? n n n n n n

Fishable biomass B P 3-yr ave is +/- 10% of previous 3-yr ave +10%

+1%

+2%

+71%

Harvest fraction B P above 32% (int. standard) 3%

4%

10%

4%

Egg production B S <20% of pristine pop 28% 20% 30%

33%

Recruitment B S 3-yr ave is +/- 10% of historical mean +73%

-50%

+90%

+198%

3-yr ave is +/- 10% of previous 6-yr ave +211%

+13%

+14%

+11%

Age composition B P prop >10 yrs <20% of fished pop 32.3% 22.0% 19.7% 10.4% 14.2% na

Page 54: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

45

4. DISCUSSION

4.1 Information for stock assessment

This assessment has considered the results from several long-term monitoring programs to

determine the current status of the stocks of snapper in South Australia. The datasets were:

commercial catch and effort data from 1984 to 2015; and data on population size and age

structures that were collected throughout the 2000s up to 2015. They were augmented by

results from three telephone/diary recreational surveys. All data were integrated in the

SnapEst computer fishery model. Time series for a suite of general and biological fishery

performance indicators were generated and the data for 2015 were compared against

trigger reference points calculated from the historical time series from 1984 to 2015, in a

weight-of-evidence approach. The outcome of the process is presented using the

classification system of Flood et al. (2014) (Table 1.3).

In this report, the spatial scale for classifying stock status differed from earlier reports

(Fowler et al. 2013), based on findings of the recently-completed, FRDC-funded project into

the stock structure of SA’s snapper population (Fowler 2016). That project considered

large-scale movement patterns and stock structure in the context of significant,

unprecedented changes in the spatial structure of this fishery and uncertainty about the

underlying demographic processes. It concluded that South Australian snapper involved

three stocks; the Spencer Gulf / West Coast Stock (SG/WC), the Gulf St. Vincent Stock

(GSV), and the cross-jurisdictional Western Victorian Stock (WVS). Each stock involves a

number of regional populations, one of which includes a significant nursery area that is the

primary source of recruitment for the whole stock. These nursery areas are located in NSG,

NGSV and Port Phillip Bay (PPB), Victoria. The populations in these natal regions are self-

replenishing, whilst adjacent populations depend on immigration from these natal regions.

Such movement may occur over distances of up to hundreds of km. The rates of emigration

vary amongst years as determined by inter-annual variation in recruitment of 0+ fish to the

nursery areas. The natal regions experience higher recruitment than do the distant regions,

as the latter depend on overflow from the former. This is likely to be a density-dependent

process that occurs at a greater rate following strong recruitment year classes. Therefore,

the demographic processes that drive population dynamics differ amongst regions.

Consequently, in this report, the trends in fishery performance indicators were considered at

the regional scale to elucidate the recent demographic processes. Then, based on this

refined understanding, the results from component regions were assessed together to

assign stock status at the scale of biological stock.

Page 55: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

46

4.2 Status of snapper stocks

4.2.1 Spencer Gulf / West Coast Stock

Northern Spencer Gulf

NSG supports a self-replenishing population based around a significant nursery area

(Fowler 2016). During the 1990s, this was the most productive region in South Australia’s

snapper fishery that consistently contributed greater than 60% of the State’s commercial

catch. However, total catches declined through the 2000s and then in 2012 dropped to

approximately 30% of that of 2011. From 2013 to 2015, total catches remained low,

producing the four lowest consecutive annual catches for the region. Understanding this

downturn in regional productivity has been challenging because of conflicting signals from

different fishery performance indicators. This is further complicated by significant

management changes that were implemented in 2012 and 2013, in response to emerging

indications of a risk to sustainability (Fowler et al. 2013, PIRSA 2013).

Targeted handline and longline effort declined in this region throughout the 2000s, with

particular drops in 2012. Declining effort is significant as fishing effort is recognised as a

primary performance indicator for snapper, thought to relate to fishable biomass (PIRSA

2013). This is because snapper is a premium species that returns a high value per unit

weight to fishers, and so should attract high levels of targeted effort as long as sufficient

biomass is available. Furthermore, the considerable drops that occurred in targeted catch

and effort in 2012, were associated with a significant decline in handline CPUE from 141 to

68 kg.fisherday-1. Whilst management changes implemented in late 2012, i.e. daily trip

limits, gear restrictions and extension of the seasonal closure, may have contributed, to

some extent, to these declines in primary and secondary performance indicators, they

cannot fully account for the declines. Rather, the declines in targeted catch, effort and

CPUE are likely to represent a considerable decrease in fishable biomass throughout the

2000s, but particularly from 2012. The elevated levels of CPUE prior to 2012 have

previously been attributed to hyperstability, consistent with the hypothesis of declining

biomass through the 2000s (Fowler and McGlennon 2011, Fowler et al. 2013). In 2014 and

2015, catch and effort remained low, but handline CPUE, as well as Prop250kgTarHL and

Prop250kgTarLL increased to some extent, suggesting some recent recovery in biomass.

It was hypothesised above that the empirical data and fishery performance indicators for

snapper in NSG in the late 2000s were consistent with declining levels of fishable biomass.

The population age structures provide a possible explanation for this trend that relates to

Page 56: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

47

inter-annual variation in recruitment rates (Fig. 6.2). From 2000 onwards, the successive

age structures show the diminishing significance of the strong 1991 year class and

emergence of the 1997 and 1999 year classes that were both significant contributors to

fishery catches until 2015. However, no year classes that recruited throughout the 2000s

contributed to fishery catches as regularly or to the same extent as those from 1997 and

1999. As such, year classes that had recruited to the population through the 1990s

accounted for relatively high proportions of fishery catches during the 2000s. Consequently,

the declining biomass through the 2000s, particularly from 2012 to 2015, reflects the

relatively weak recruitment year classes throughout the 2000s.

In contrast to the interpretation of empirical data, the model-estimated biomass for NSG

demonstrated an increasing trend from 1994, which accelerated in 2014 and 2015. As

such, there was a discrepancy between the interpretation of the empirical data and the

outputs of the model. This reflects different emphasis being placed on the fishery statistics

and population age structures. The interpretation of the empirical data considered that the

relatively weak year classes from 2000 onwards indicated poor recruitment. Conversely,

the SnapEst model focussed on the persistence of 10 – 20 year old fish in the population,

reflective of low rates of mortality further compounded by recent low and declining levels of

fishing effort that depressed model-estimated mortality and exploitation rates. Such low

exploitation of the stock was interpreted by the model as “adult fish being left in the water,

providing opportunity for the population to increase through reproduction and recruitment”.

This discrepancy is considered further in Section 4.5, Future research needs.

Southern Spencer Gulf

Through the 2000s, this region supported a highly productive snapper fishery which had two

peak periods in production. The more significant peak was between 2005 and 2009, when

SSG produced the highest annual regional catches for the State, surpassing those of NSG.

This reflected a strong handline fishery that was augmented by development of the longline

sector, associated with the uptake by fishers of new longline fishing technology.

Nevertheless, after the peak productivity of the fishery in 2007 when this region contributed

54.0% of the State-wide catch, targeted handline and longline catches, effort and CPUE all

declined markedly to 2015, some to their lowest levels. In 2014 and 2015, the region

contributed only 3.5% and 4.7%, respectively of the State’s total catches. In 2015, the low

estimates of fishery catch, effort and CPUE culminated in two negative breaches of general

performance indicators (i.e. for total catch and targeted handline effort). As these declining

trends had commenced prior to late 2012, they are unlikely to have been solely attributable

Page 57: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

48

to changes in fishery management. Rather, they are consistent with a decline in biomass

between 2007 and 2015.

The annual age structures for SSG through the 2000s were typically dominated by either

one or two strong year classes. This reflects the dependence of this region on emigration

from NSG, from where only the strongest year classes are likely to have provided significant

numbers of immigrants to SSG. The age structures for SSG up to 2014 were dominated by

the 1991, 1997 and 1999 year classes. The former year class largely accounted for the

peak in catch around 2001, whilst the latter two contributed to the larger peak around 2007.

In comparison, recruitment throughout the 2000s was poor. From 2007, the declining

catches in SSG reflected the lack of strong year classes subsequent to the 1999 year class,

whilst the older year classes were gradually depleted through fishing and natural mortality.

As with NSG, for this region there was a discrepancy in biomass between the interpretation

of the empirical data and the estimates from the model. SnapEst estimates support the

hypothesis of low recruitment into SSG subsequent to the 1997 and 1999 year classes,

which resulted in a decline in biomass after 2006. Nevertheless, due to lower mortality rates

inferred from the declining commercial fishing effort and model-estimated harvest fractions,

the extent of the decline in model-estimated biomass was modest.

West Coast

This region is considered part of the SG/WC Stock based on evidence of connectivity from

an early otolith chemistry study (Fowler et al. 2005b). It is considered that when strong year

classes recruited to NSG, some fish emigrated from SG to the WC (Fowler 2016). There

are no recent age structures for this region because of the difficulty in accessing fishery

catches to obtain biological samples. As such, the only fishery performance indicators that

can be considered are commercial fishery statistics. The region has historically produced

small catches relative to the gulfs. In 2008 and 2009, relatively high levels of targeted

handline and longline catch and effort culminated in record catches and catch rates. Since

then, total catch has decreased reflecting declining levels of targeted catch, effort and

CPUE. This region produced 7% of the State’s catch in 2008, which was halved by 2012

and has since remained low. Such declines culminated in activation of two trigger reference

points in 2015, including the 3rd lowest handline effort yet recorded. As declining trends in

fishery statistics had commenced prior to the management changes in late 2012, they

suggest a declining level of biomass. This is consistent with minimal emigration from NSG,

reflecting the lack of strong recruitment year classes to the nursery area through the 2000s.

Page 58: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

49

Assignment of Stock Status

The assignment of stock status for the SG/WC Stock has been challenging because of

conflicting signals from different fishery performance indicators, particularly with respect to

recent trends in fishable biomass. The empirical data were interpreted as reflecting

declining levels of biomass as a consequence of poor recruitment to NSG through the

2000s. Alternatively, the time-series of fishable biomass from the SnapEst model increased

in both NSG and SSG. Nevertheless, there was some consistency in the two interpretations

as both indicated poor recruitment through the 2000s, relative to the 1990s. Anecdotal

reports from the fishing industry suggest that the stock in SG is in a poor state, relative to

periods of record productivity during the early and mid-2000s. As such, the model estimates

of biomass for both regions appear to be positively biased. Nevertheless, the model outputs

have usefully drawn attention to the low exploitation rates and low levels of fishing effort.

The model output also indicated that recruitment in NSG increased in 2014 and 2015,

reflecting the recent upturn in handline CPUE.

Whilst the biomass of the SG/WC has declined and recruitment was below average through

the 2000s, there is some evidence of recent recruitment. Low fishing effort in the three

regions indicates that exploitation rates have fallen to low levels. Furthermore, significant

management changes were implemented in 2012 and 2013 that were focussed on reducing

commercial catch and maximising reproductive output and recruitment. On the basis of the

evidence provided above, the SG/WC Stock is classified as transitional depleting.

4.2.2 Gulf St. Vincent Stock

Northern Gulf St. Vincent

The region of NGSV includes one of the primary nursery areas for snapper in south eastern

Australia, and is a self-replenishing population (Fowler 2016). Historically, it supported only

very low catches through the 1980s, 1990s and early 2000s. However, from 2007 to 2010,

total catch increased to not only a regional peak but this region became the dominant

contributor to the State-wide catch, whose contributions increased from 9% in 2007 to 74%

in 2015. High longline catches related to an exponential increase in longline effort and a

substantial increase in longline catch rate. The increase in effort was, in part, due to a

substantial influx of longline fishers to the region. Contemporaneously, there were marginal

increases in targeted handline effort, catch and CPUE. Whilst the large increases for this

region are likely to have partly been related to the greater efficiency of the new longline

fishing gear, the data are also consistent with a substantial increase in fishable biomass

through the 2000s, relative to earlier years.

Page 59: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

50

The recent population age structures are informative about the recruitment history to the

region. They demonstrate that several year classes through the 2000s (i.e. 2001, 2004,

2006, 2007 and 2009), each made significant contributions to catch and were persistent in

time. Consequently, the high catches from 2010 onwards involved a high proportion of fish

that had recruited to the population through the 2000s. In this sense it contrasts

considerably with SG for which the 1997 and 1999 year classes remained significant

contributors to the regional populations.

For NGSV, the trends in model-estimated biomass increased considerably up to 2010,

relating to strong, model-estimated year classes throughout the 2000s. As such, the model

outputs are consistent with the interpretation of fishery statistics presented above.

Furthermore, the estimated exploitation rates increased substantially over time, reflecting

the substantial increase in effort that occurred.

Southern Gulf St. Vincent

SGSV is the most demographically complex of the six South Australian regional populations

(Fowler 2016). It is located at the boundary of, and in the mixing zone between the three

stocks, and likely to receive some supplementation from each. Nevertheless, such

supplementation appears to be low, as this region has never produced the substantial

catches that the other gulf regions have at different times. In every year since 1994, SGSV

has produced <6% of the State’s total catch. From 2009 to 2011, longline catches and

catch rates increased marginally, resulting in higher total catches than taken during the

early 2000s. These increases were concomitant with those in NGSV and the SE, but

nevertheless on a smaller scale. In 2011, when this region produced its highest catch since

1990, it accounted for only 4.0% of the State’s total catch. This peak was associated with

expansion of the longline sector, which occurred contemporaneously with those in NGSV

and the SE. Since then, total catch has declined, reflecting decreases in longline catch,

effort, CPUE and number of fishers. In 2015, one general trigger reference point was

breached along with one each for Prop250kgTarHL and Prop250kgTarLL.

Overall, the fishery statistics suggest that this region supports a relatively low biomass of

snapper, which has experienced marginal decline since 2011. In the age structures, some

year classes were stronger than others including those in 1991, 1997, 2001, 2004, 2006,

2007 and 2009. Recent catches were dominated by year classes that recruited after 2000.

The complex age structures reflect the influence of immigration from different sources. The

2001 fish had originated in PPB and were part of the WVS, whilst the 1991, 1997 and 2004

Page 60: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

51

year classes had originated in the northern gulfs (Fowler 2016). The origins of the 2006,

2007 and 2009 year classes remain undetermined.

The model output for this region indicated an increasing biomass reflecting four strong

recruitment year classes through the 2000s, as well as low exploitation rates. The

comparison of model and data Z (total mortality) for SGSV show a clear difference. The

model estimates were lower, implying that the model has under-estimated mortality rate for

this region compared to that estimated from raw age samples, suggesting that the model

has over-estimated biomass. The demographic complexity of this population and likely

migration into it in recent years pose challenges to the underlying assumptions of the

SnapEst model.

Assignment of Stock Status

Since the mid-2000s, the bulk of the biomass for the GSV Stock has been located in the

northern part of the gulf. There was considerable consistency in the general and biological

fishery performance indicators for this region that reflected a substantial increase in biomass

relating to a succession of strong year classes. In 2015, there were marginal declines in

total catch and CPUE that may have related to management changes in 2012 and 2013, or

to the marginal declines in biomass due to the long-term high effort. Nevertheless, there

were no indications that the biomass was moving in the direction of being overfished. For

SGSV, the model is likely to have provided overly-optimistic estimates of biological

performance indicators. Nevertheless, the recent trends in general fishery performance

indicators were relatively stable.

The above evidence indicates that the biomass of this stock is unlikely to be recruitment

overfished and that the current level of fishing mortality is unlikely to cause the biomass to

become recruitment overfished. On this basis the GSV Stock is classified as sustainable.

However, based on the extent and speed of the decline of fishery status in SG that was, to

some extent, hidden through the maintenance of high catch rates through hyperstability, it is

possible that similar processes could occur in NGSV.

4.2.3 Western Victorian Stock

South East

The recent analysis of adult movement patterns and stock structure (Fowler 2016),

determined that the SE regional population is part of the WVS. This stock depends on

recruitment into the primary nursery area of PPB, Victoria. From there, young fish emigrate

Page 61: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

52

and subsequently move and replenish the populations along the west coast of Victoria

(Hamer et al. 2005, 2006, 2011), and south eastern South Australia (Fowler 2016).

Historically, the snapper fishery in the SE was largely incidental, producing low catches,

associated with low effort and catch rates. However, it experienced a dramatic, exponential

increase in total catch from 2005, which peaked at 261 t in 2010. In that year, this region

contributed 25% of the State’s total commercial catch. Nevertheless, since then, the

regional catch has fallen back towards low historic levels. The unprecedented increase in

catch was primarily driven by significant increases in targeted catch, effort, CPUE and

number of fishers in the longline sector. This period was contemporaneous with the

developing longline sector in NGSV, and is probably partly attributable to the adoption of

similar new fishing practices. However, the expansion also suggests a significant,

unprecedented, episodic increase in biomass of snapper, which has subsequently declined.

Population age structures indicate that the episodic increase related to recruitment of strong

2001 and 2004 year classes into the population. These were the two strongest year classes

to recruit to PPB in the 24 years from 1993 to 2016 (Hamer and Conron 2016). As such, the

population has been dominated by fish that recruited since 2000.

As the SE regional population is part of the WVS, its stock status largely depends on that of

the main reproductive population in and around PPB, Victoria. A stock assessment was

recently completed in Victoria for the WVS (Hamer and Conron 2016). The stock was

assigned the status of ‘sustainable’, largely based on results of the 0+ recruitment survey,

which showed above-average recruitment levels for six of the past 12 years. As such, the

SE regional population is classified as sustainable. Supplementation of the SE region is

already apparent with several new year-classes apparent in recent age structures.

4.3 Conclusions

This stock assessment was the beneficiary of the project on adult snapper movement and

stock structure (Fowler 2016), which informed about the obligate connectivity between

different regions through fish movement and the consequences for regional population

dynamics. The fishable biomass in both SSG and the WC depend on emigration from

NSG. That in the SE depends on movement over hundreds of km from PPB. For the

source regions that contain the primary nursery areas in the gulfs (i.e. NSG and NGSV), the

population dynamics are driven by inter-annual variability in recruitment. For the destination

regions of SSG, SGSV, SE and WC, the population dynamics are driven by emigration from

the source regions.

Page 62: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

53

In this assessment, the fishery statistics, population age structures and model outputs were

considered at the regional scale, and subsequently interpreted at the scale of biological

stock (Fowler 2016). The SG/WC Stock is classified as transitional depleting, reflecting

the declining levels of regional biomass as a consequence of relatively poor recruitment into

NSG through the 2000s. As no strong year classes have recruited to this region since

1999, there has been minimal emigration of fish to SSG and WC. In contrast, the SE region

was the beneficiary of two strong year classes that recruited to PPB, and subsequent

emigration. Those year classes have now been depleted and the biomass has declined to a

lower level. Nevertheless, as there has been above-average recruitment into PPB for six of

the past 12 years (Hamer and Conron 2016), emigration to the SE is expected to continue.

As such, this stock is classified as sustainable. The GSV Stock experienced an

unprecedented increase in biomass through the 2000s based on numerous strong

recruitment year classes to NGSV. Consequently, this stock is considered sustainable

despite recent small declines for some indicators. The poor status of the SG/WC Stock

provides a warning about NGSV with respect to the impact of sustained high fishing effort.

4.4 Uncertainties in the assessment

Our primary uncertainty about this assessment relates to the discrepancy between trends in

biomass of snapper from the empirical data and output from the fishery model SnapEst.

This eventuated from the emphasis placed on different datasets. Declining trends in fishery

catch, effort and CPUE and recent low year class strength were interpreted in terms of

declining biomass for the SG/WC Stock. Alternatively, the SnapEst model interpreted broad

age structures and low levels of fishing effort in terms of low exploitation rates, leading to

increasing fishable biomass. This conflict reflects that the fishery performance indicators,

such as CPUE and fishing effort, used in this fishery are not necessarily closely related to

fishable biomass. Furthermore, because of the multi-species nature of this fishery, such

parameters are not only influenced by changes in fishable biomass, but by other factors

such as fisher behaviour, as fishers can move between regions or species to pursue better

financial gain.

There are several further concerns about using commercial fishery statistics as indicators of

stock biomass. Since 1984, when the collection of detailed data commenced for the

commercial sector in SA, there has been considerable technological creep in this industry.

This includes: the use of electronic equipment for finding fish and navigation; the increase in

power and range of vessels; and improvements to fishing technology, such as with the

adoption in SA of new monofilament longline gear during the mid-2000s. The influences of

these developments on ‘effective’ effort are unknown. In this assessment, a statistical

Page 63: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

54

approach was used to de-trend the time-series of CPUE, i.e. to remove long-term trends

that could relate to increasing ‘effective’ effort. Such de-trending is useful if biomass has not

increased over time, but is less useful when there has been a considerable change in

biomass, as occurred for NGSV during the 2000s. This means that differentiating between

time-series of CPUE and their de-trended versions is complex and requires subjective

interpretation. This situation is not yet resolved for snapper in SA.

The second concern regarding the use of commercial fishery statistics as indicators of

biomass relates to the management arrangements that were introduced in 2012 and 2013,

which included a daily catch limit of 500 kg and reduction in number of longline hooks.

Besides limiting catches and thereby impacting on estimates of regional catch and CPUE,

these changes are also likely to have impacted on the usefulness of CPUE as an indicator

of biomass. This was the reason for developing the two new fishery performance indicators

of ‘Prop250kgTarHL’ and ‘Prop250kgTarLL’ (PIRSA 2013), i.e. the proportions of catches

taken per gear type per year (excluding November to January) that reach 250 kg. They

were developed to replace CPUE as a measure of relative abundance. Nevertheless,

between 2004 and 2015, both parameters were highly variable from year-to-year, making

questionable their relationship with biomass and value as indicators.

A further key uncertainty for South Australia’s snapper fishery remains the poor

understanding of recreational catch and effort. Uncertainties about recreational catch and

effort remain considerable, whilst the spatial and temporal resolution of the data is poor.

The results from the three State-wide telephone / diary surveys that were undertaken

throughout the 2000s, have consistently indicated that the recreational sector accounts for

considerable proportions of total catch, but the estimates have varied considerably. Such

variability provides challenges for fishery modelling, stock assessments and fishery

management.

4.5 Future research needs

The discrepancy in interpretation of fishery performance indicators for snapper, particularly

for NSG, has highlighted the current dependence on fishery-related data. Currently, there is

no fishery performance indicator that directly relates to fishable biomass or recruitment, as

is the case for Victoria (Hamer and Conron 2016). Consequently, different performance

indicators can provide ambiguous outcomes with respect to stock status. Awareness of this

issue several years ago highlighted the need for a fishery-independent estimate of biomass

for snapper. FRDC Project 2014/019 has assessed the Daily Egg Production Method for

Page 64: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

55

estimating the biomass of snapper in SA. This project is currently nearing completion, and

the outcomes should provide an important new source of data that will be a useful empirical

anchor point for the SnapEst model. Also, a review of the Management Plan that is

scheduled for 2018 will provide an opportunity for assessing the usefulness of the current

suite of performance indicators, including Prop250kgTarHL and Prop250kgTarLL.

Part of the confusion in interpreting fishery performance indicators in this assessment

related to the inconsistency between some populations reputedly having low levels of

biomass, but at the same time still including 10 – 20 year old fish. It would be expected that

if fishing mortality was high enough to reduce biomass, then the population would have

experienced age truncation and old fish would not remain in the population. It is unknown

whether through their behaviour some fish gain access to a refuge from fishing. As such,

our understanding of snapper behaviour remains incomplete, despite recent advancements

in this area (Fowler 2016). This issue relates to habitat use and movement at the local

scale and would be appropriately addressed using acoustic telemetry.

Page 65: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

56

5. REFERENCE LIST

Crossland J (1976). Snapper tagging in north-east New Zealand, 1974: analysis of methods, return

rates and movements. New Zealand Journal of Marine and Freshwater Research 10: 675-686. Donnellan SC, McGlennon D (1996). Stock identification and discrimination in snapper (Pagrus

auratus) in southern Australia. Final Report to the Fisheries Research and Development Corporation, 31 pp.

Dunn, A, Francis RICC, Doonan IJ (2002). Comparison of the Chapman–Robson and regression

estimators of Z from catch-curve data when non-sampling stochastic error is present. Fisheries Research 59: 149-159.

Flood M, Stobutzki I, Andrews J, Ashby C, Begg G, Fletcher R, Gardner C, Georgeson L, Hansen S,

Hartmann K, Hone P, Horvat P, Maloney L, McDonald B, Moore A, Roelofs A, Sainsbury K, Saunders T, Smith T, Stewardson C, Stewart J, Wise B (eds) (2014). Status of key Australian fish stocks reports 2014, Fisheries Research and Development Corporation, Canberra.

Fowler AJ (2000). Snapper (Pagrus auratus). South Australian Fisheries Assessment Series 2000/13.

55 pp. Fowler AJ (2002). Snapper (Pagrus auratus). South Australian Fisheries Assessment Series 2001/13.

64 pp. Fowler (2016). The influence of fish movement on regional fishery production and stock structure for

South Australia’s snapper (Chrysophrys auratus) fishery. Final Report to FRDC (Project No. 2012/020). 181 pp.

Fowler AJ, Jennings PR (2003). Dynamics in 0+ recruitment and early life history for snapper (Pagrus

auratus, Sparidae) in South Australia. Marine and Freshwater Research 54: 941-956. Fowler AJ, McGlennon D (2011). Variation in productivity of a key snapper, Chrysophrys auratus,

fishery related to recruitment and fleet dynamics. Fisheries Management and Ecology 18: 411-423.

Fowler AJ, McGarvey R, Feenstra JE, Jackson WB, and Jennings PR (2003). Snapper (Pagrus

auratus) Fishery. Fishery Assessment Report to PIRSA for the Marine Scalefish Management Committee. 70 pp.

Fowler AJ, Gillanders BM, Hall KC (2005b). Relationship between elemental concentration and age

from otoliths of adult snapper (Pagrus auratus, Sparidae): implications for movement and stock structure. Marine and Freshwater Research 56: 661-676.

Fowler AJ, McGarvey R, Feenstra JE, Jackson WB, and Jennings PR (2005a). Snapper (Pagrus

auratus) Fishery. Fishery Assessment Report to PIRSA for the Marine Scalefish Management Committee. South Australian Research and Development Institute (Aquatic Sciences), Adelaide, RD03/0068-2. 99 pp.

Fowler AJ, McGarvey R, Feenstra JE and Jackson WB (2007). Snapper (Pagrus auratus) Fishery.

Fishery Assessment Report to PIRSA Fisheries. South Australian Research and Development Institute (Aquatic Sciences), Adelaide, F2007/000523-1, SARDI Research Report Series No. 224. Pp 90.

Fowler AJ, McGarvey R, Burch P, Feenstra JE and Jackson WB (2010). Snapper (Chrysophrys

auratus) Fishery. Fishery Assessment Report to PIRSA Fisheries and Aquaculture. South Australian Research and Development Institute (Aquatic Sciences), Adelaide, F2007/000523-2, SARDI Research Report Series No. 473. Pp 109.

Page 66: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

57

Fowler AJ, McGarvey R, Burch P, Feenstra JE, Jackson WB, Lloyd MT (2013). Snapper (Chrysophrys auratus) Fishery. Fishery Assessment Report to PIRSA Fisheries and Aquaculture. South Australian Research and Development Institute (Aquatic Sciences), Adelaide, F2007/000523-3, SARDI Research Report Series No. 713. Pp 103.

Fukuhara O (1985). Functional morphology and behaviour of early life stages of red sea bream.

Bulletin of the Japanese Society of Scientific Fisheries 51: 731-743. Gilbert DJ, Davies NM, McKenzie JR (2006) Development of an age–length structured model of the

Hauraki Gulf–Bay of Plenty snapper (Pagrus auratus) population. Marine and Freshwater Research 57: 553–568.

Giri K, Hall K (2015) South Australian Recreational Fishing Survey. Fisheries Victoria Internal Report

Series No. 62. Hamer P, Conron S (2016). Snapper Stock Assessment 2016. Fisheries Victoria Science Report

Series No. 10, Fisheries Victoria, Queenscliff. Hamer PA, Acevedo S, Jenkins GP, Newman A. (2011). Connectivity of a large embayment and

coastal fishery: spawning aggregations in one bay source local and broad-scale fishery replenishment. Journal of Fish Biology 78: 1090-1109.

Hamer PA, Jenkins GP, Coutin P (2006). Barium variation in Pagrus auratus (Sparidae) otoliths: a

potential indicator of migration between an embayment and ocean waters in south-eastern Australia. Estuarine, Coastal and Shelf Science 68: 686-702.

Hamer PA, Jenkins GP, Gillanders BM (2005). Chemical tags in otoliths indicate the importance of

local and distant settlement areas to populations of a temperate sparid, Pagrus auratus. Canadian Journal of Fisheries and Aquatic Sciences 62: 623-630.

Henry GW, Lyle JM (2003). The National Recreational and Indigenous Fishing Survey. Final Report

to FRDC (Project No. 1999/158). NSW Fisheries Final Report Series no. 48. 188 pp. Jackson G, Fowler A, Holmes B, Kemp J, Stewart J (2012). Snapper Pagrus auratus in M Flood, I

Stobutzki, J Andrews, G Begg, W Fletcher, C Gardner, J Kemp, A Moore, A O’Brien, R Quinn, J Roach, K Rowling, K Sainsbury, T Saunders, T Ward & M Winning (eds), Status of key Australian fish stocks reports 2012, Fisheries Research and Development Corporation, Canberra, 344-354.

Jones GK (2009). South Australian Recreational Fishing Survey. PIRSA Fisheries, Adelaide, 84 pp.

South Australian Fisheries Management Series Paper No. 54. Jones GK, Doonan A (2005). 2000-01 National Recreational and Indigenous Fishing Survey – SA

Regional information. South Australian Fisheries Management Series Paper no. 46. 99 pp. Kailola PJ, Williams MJ, Stewart PC, Reichelt RE, McNee A, Grieve C (1993). Australian Fisheries

Resources. Bureau of Resource Sciences and the Fisheries Research and Development Corporation Canberra Australia. Imprint Limited, Brisbane.

Last PR, White WT, Gledhill DC, Hobday AJ, Brown R, Edgar GJ, Pecl G (2011). Long-term shifts in

abundance and distribution of a temperate fish fauna: a response to climate change and fishery practises. Global ecology and Biogeography 20: 58-72.

Matsuyama M, Adachi S, Nagahama Y, Matsuura S (1988). Diurnal rhythm of oocyte development

and plasma steroid hormone levels in the female Red Sea Bream, Pagrus major, during the spawning season. Aquaculture 73: 357-372.

McGarvey R (2004). Model comparison of seasonal closure and maximum size for management of the

South Australian snapper fishery. Ecological Management & Restoration 5: 139-141.

Page 67: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

58

McGarvey R, Jones GK (2000). Biological requirements and yield- and egg-per-recruit estimates for management of the South Australian snapper (Pagrus auratus) fishery. Report to the Marine Scalefish Fishery Management Committee.

McGarvey R, Feenstra JE (2004). Stock assessment models with graphical user interfaces for key

South Australian marine finfish stocks. Final Report to FRDC (Project No. 1999/145). South Australian Research and Development Institute (Aquatic Sciences), Adelaide. 176 pp.

McGarvey R, Feenstra JE, Ye Q (2007). Modeling fish numbers dynamically by age and length:

partitioning cohorts into ‘slices’. Canadian Journal of Fisheries and Aquatic Sciences 64: 1157-1173.

McGarvey R, Fowler AJ (2002). Seasonal growth of King George whiting (Sillaginodes punctata) from

length-at-age samples truncated by legal minimum size. Fishery Bulletin 100: 545-558. McGarvey R, Burch P, Feenstra JE (2010). Management Strategy Evaluation for South Australian

Snapper. SARDI Aquatic Sciences Publication No. F2010/00026-1. SARDI Research Report Series No. 441. April 2010. 32 pp.

McGarvey R., Burch P, Feenstra JE (2011). Use of management strategy comparison scatterplots to

identify clusters of high performing strategies. Fisheries Management and Ecology 18: 349–359.

McGlennon D (1999). Snapper (Pagrus auratus). South Australian Fisheries Assessment Series 98/7. McGlennon D (2004). The fisheries biology and population dynamics of snapper Pagrus auratus in

northern Spencer Gulf, South Australia. PhD thesis, University of Adelaide. 212 pp. McGlennon D, Jones GK (1997). Snapper (Pagrus auratus). South Australian Fisheries Assessment

Series 97/07. 59 pp. McGlennon D, Jones GK (1999). Snapper (Pagrus auratus). South Australian Fisheries Assessment

Series 99/13. 23 pp. McGlennon D, Jones GK, Baker J, Jackson WB, Kinloch MA (2000). Ageing, catch-at-age and relative

year-class strength for snapper (Pagrus auratus) in Northern Spencer Gulf, South Australia. Marine and Freshwater Research 51: 669-677.

McGlennon D, Kinloch MA (1997). Resource allocation in the South Australian marine scalefish

fishery. FRDC project 93/249, Final Report. 105 pp. Moran M, Burton C, Caputi N (1998). Sexual and local variation in head morphology of snapper,

Pagrus auratus, Sparidae, in the Shark Bay region of Western Australia. Marine and Freshwater Research 50: 27-34.

Moran M, Burton C, Jenke J (2003). Long-term movement patterns of continental shelf and inner gulf

snapper (Pagrus auratus, Sparidae) from tagging in the Shark Bay region of Western Australia. Marine and Freshwater Research 54: 913-922.

Murphy HM, Jenkins GP, Hamer PA, Swearer S (2012). Interannual variation in larval survival of

snapper (Chrysophrys auratus, Sparidae) is linked to diet breadth and prey availability. Canadian Journal of Fisheries and Aquatic Sciences 69: 1340-1351.

Murphy HM, Jenkins GP, Hamer PA, Swearer S (2013). Interannual variation in larval abundance and

growth in snapper, Chrysophrys auratus, Sparidae, related to prey availability and temperature. Marine Ecology progress Series 487: 151-162.

Norriss JV, Crisafulli B (2010). Longevity in Australian snapper Pagrus auratus (Sparidae). Journal of

the Royal Society of Western Australia 93: 129-132.

Page 68: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

59

Paul LJ (1967). An evaluation of tagging experiments on the New Zealand snapper, Chrysophrys auratus (Forster), during the period of 1952 to 1963. New Zealand Journal of Marine and Freshwater Research 1: 455-463.

PIRSA (2013). Management Plan for the South Australian Commercial Marine Scalefish Fishery.

PIRSA Fisheries and Aquaculture, Adelaide, 143 pp. The South Australian Fishery Management Series, Paper No. 59.

PIRSA (2016a). Review of size, bag and boat limits in South Australia’s recreational fishing sector,

marine and freshwater. Support document for the management plan for recreational fishing in South Australia. January 2016.

PIRSA (2016b). Draft Management Plan for Recreational Fishing in South Australia. Public

Consultation Document. January 2016. Sanders MJ (1974). Tagging indicates at least two stocks of snapper Chrysophrys auratus in south-

east Australian waters. New Zealand Journal of Marine and Freshwater Research 8: 371-374. Saunders RJ (2009). The reproductive biology and recruitment dynamics of snapper, Chrysophrys

auratus. PhD thesis, University of Adelaide. 154 pp. Saunders RJ, Fowler AJ, Gillanders BM (2012). The spawning dynamics of snapper (Chrysophrys

auratus) in northern Spencer Gulf, South Australia. New Zealand Journal of Marine and Freshwater Research 46: 491-510.

Scott SG, Zeldis JR, Pankhurst NW (1993). Evidence of daily spawning in natural populations of the

New Zealand snapper Pagrus auratus (Sparidae). Environmental Biology of Fishes 36: 149-156.

Steer, M.A., McGarvey, R., Carroll, J., Jackson, W. B., Lloyd, M. T. and Feenstra, J. E. (2016). Southern

Garfish (Hyporhamphus melanochir) Fishery. Fishery Assessment Report to PIRSA Fisheries and Aquaculture. South Australian Research and Development Institute (Aquatic Sciences), Adelaide. SARDI Publication No. F2007/000720-4. SARDI Research Report Series No. 891. 75pp.

Tanaka M (1985). Factors affecting the inshore migration of pelagic larval and demersal juvenile red

sea bream Pagrus major to a nursery ground. Transactions of the American Society 114: 471-477.

Thorpe RS (1975). Quantitative handling of characters useful in snake systematics with particular

reference to intraspecific variation in the ringed snake Natrix natrix (L.), Biological Journal of the Linnean Society 7: 27-43.

Zeldis JR, Oldman J, Ballara S, Richards LA (2005). Physical fluxes, pelagic ecosystem structure, and

larval fish survival in Hauraki Gulf, New Zealand. Canadian Journal of Fisheries and Aquatic Sciences 62: 593-610.

Page 69: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

60

6. APPENDICES

6.1 Annual size and age structures

This section presents the annual estimates of population size and age structures

determined from SARDI’s market sampling program that has focussed on Adelaide’s

SAFCOL fish market. The data available in the different years varied amongst regions,

based on the availability of captured fish from the commercial sector, as determined by

variation in regional productivity.

Page 70: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

61

Fig. 6.1 Size and biomass distributions for snapper caught in NSG in each of 13 years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class.

Page 71: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

62

Fig. 6.2 Estimated age structures of fish caught in NSG in each of 13 years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned.

Page 72: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

63

Fig. 6.3 Size and biomass distributions for snapper caught in SSG in each of 12 years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class.

Page 73: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

64

Fig. 6.4 Estimated age structures of fish caught in SSG in each of 12 years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned.

Page 74: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

65

Fig. 6.5 Size and biomass distributions for snapper caught in NGSV in each of eight years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class.

Page 75: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

66

Fig. 6.6 Estimated age structures of fish caught in NGSV in each of seven years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned.

Page 76: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

67

Fig. 6.7 Size and biomass distributions for snapper caught in SGSV in each of 12 years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class.

Page 77: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

68

Fig. 6.8 Estimated age structures of fish caught in SGSV in each of seven years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned.

Page 78: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

69

Fig. 6.9 Size and biomass distributions for snapper caught in SE in each of seven years. Left hand graphs show the size structures. Right hand graphs show the percentage of biomass accounted for by each size class.

Page 79: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

70

Fig. 6.10 Estimated age structures of fish caught in SE in each of six years. For each year, data are presented as the relative percentage of total catch accounted for by each year class, i.e. the years in which they were spawned.

Page 80: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

71

6.2 Snapper stock assessment model

In this section we summarise the following components of the stock assessment model: (1)

growth, (2) recruitment, (3) the population array including length slices, (4) mortality, and (5)

the likelihood function relating model to data. The slice-partition method, with detailed

pseudo-code, is described in Appendix C of the 2015 Garfish stock assessment report

(Steer et al. 2016).

6.2.1 Growth

The starting point and basis of the slice method for partitioning fish cohorts by length is the

length-at-age growth submodel. A statistical growth submodel is needed which fully specifies

the probability density function (pdf) of fish lengths for each model age. This represents the

(normal) distribution of fish by length in each cohort age that would be observed in the

absence of length-asymmetric mortality, because length-selective capture mortality will

subsequently be imposed on these model cohorts, after they are partitioned into slices. To

model mean fish length l , the mean of the normal length-at-age pdf, for any half-yearly cohort

age, a , we employed a 4-parameter exponent-generalized von Bertalanffy mean length-at-

age curve: 0)( 1 exp

2

r

a tl a L K

(McGarvey and Fowler 2002). Using

two additional parameters, the dependence of the length-at-age standard deviation ( )a is

modelled as an allometric function of mean length: 1

0( ) ( )a l a

.

The growth parameters can be estimated by fitting to length-at-age samples (1) previous to,

or (2) by integrating growth estimation into, the stock assessment likelihood. We undertook

both in that order. First we fitted the growth submodel directly to catch lengths-at-age to

obtain approximate growth parameter estimates. A likelihood probability of observation

truncated at LML was assumed to make explicit the absence of sublegal snapper in these

catch samples (McGarvey and Fowler 2002). A second growth estimation was integrated into

the stock assessment likelihood, re-estimating the two parameters that most directly

determine the mean rate of growth and spread of lengths at each age, von Bertalanffy K and

the normal length-at-age standard deviation coefficient 0 .

Starting from this growth submodel, an algorithm (described in Appendix C of Steer et al.

2016) was devised to effectively ‘slice off’ the length subintervals of fish which have grown

past legal minimum length (LML) in each model time step. Once this population number is

assigned to each newly created slice bin by transferring these fish from the sublegal

component, there is no subsequent further exchange of fish between length bins. Fish within

Page 81: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

72

slices incur only mortality. The simplification of neglecting growth diffusion among length bins

affords the slice approach large reductions in computation time compared with, for example,

a length-transition approach, which requires 2( )Ln growth-transition multiplications in each

model time step and for each cohort, where Ln is the number of length bins. In a slice partition

model, growth is quantified as the increasing length range with age of each slice subinterval,

and no computation is needed to shift fish among bins.

6.2.2 Recruitment

Recruitment is defined as the creation of the (normal) length-at-age cohort at age ab = 5 half-

years (at age 2 years) when the fastest growing fish first reach legal size. The number of fish

in each cohort at the birth age, ab, is the model estimate of yearly recruitment. Each yearly

recruit number is a freely estimated model parameter. The numbers of snapper above legal

minimum length at age ab (in the upper tail of the length at age pdf) are computed (Appendix

C of Steer et al. 2016) and defined as the first newly created slice. In subsequent model time

steps, new slices are created as the calculated proportion of sublegal fish in each cohort that

have grown into legal size since the previous time step, thereby modelling the gradual

recruitment of each cohort to fishable sizes over the number of model time steps required, as

determined by the growth submodel (Appendix C of Steer et al. 2016).

6.2.3 Model population array

The model snapper population array , , ,N t r c s is 4-dimensional, fish numbers broken

down by (1) half-yearly model time step (1 = summer 1983 to 64 = winter 2015) “ t ”, (2)

spatial region (1 = SSG, 2 = NSG, 3 = SGSV, 4 = NGSV) “ r ”,(3) cohort (i.e. year-class,

given by year of spawning) “ c ”, and (4) slice “ s ”.

Variable subscripts for winter or summer half-year ( seasont ), and cohort age in half-years ( a ),

were calculated as functions of model time step, t , and cohort year, c . Ages ran from ab to

48+ half-years, the oldest age being a 'plus' group. Snapper catch and effort, for data and

model, were divided into six effort types, Ei : (i) handline (all target types), (ii) longline (all

target types) , (iii) hauling nets and minor gears (all target types), (iv) all other commercial

gears combined , (v) charter boat, and (vi) recreational. The three commercial gears, g ,

are handline, longline, and hauling net, with handline having age selectivity modelled by a

decreasing logistic function and longline having length selectivity modelled by an increasing

logistic function. Data quantities, such as reported effort E , are denoted by a tilde.

Page 82: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

73

6.2.4 Mortality

Mortality is differentiated for legal and sublegal fish. Legal-size fish, partitioned into length

slices, are subject to both fishing and natural mortality. Length-dependent gear selectivity,

and any other length-dependent mortality processes, are applied to the length-partitioned fish

numbers, specifically in the legal size range. In addition to the knife-edge cut-off below legal

minimum length, gear-specific length selectivity is modelled for legal size snapper. Sublegal

population numbers (fish below the legal size limit) incur only natural mortality.

The catch equations were effort conditioned. Thus, fishing mortality was written as a linear

proportion of reported fishing effort for each component of catch:

(B.1) ( , , , , ) ( , , ) ( , , ) ( , ( ), ) ( , ( ), )E season E E len E age EF t r c s i q r t i E t r i s r g i s s r g i ageHY .

The catchability, q , was assumed to vary with region, season, and effort type. Length

selectivity, lens , by region for longline only, followed a logistic function of fish length, the

latter specified by the midpoint of each slice

(B.2)

501 exp , ( ) ( ) , ( )( , ( ), ) , ( ) 1

, ( ) / (1 , ( ) )

sel E E

len E sel E

sel E sel E

r r g i l s l r g is r g i s fl r g i

fl r g i fl r g i

where , ( )sel Er r g i is the logistic slope parameter, 50 , ( )El r g i is the logistic 50% level

parameter, and , ( )sel Efl r g i is a parameter to raise the level of the function over the whole

range of lengths.

A decreasing logistic selectivity function of fish age was applied for handline only, by region,

as

(B.3)

501 exp , ( ) , ( )( , ( ), ) 1 1

, ( ) / (1 , ( ) )

agesel E agesel E

age E

agesel E agesel E

r r g i ageHY l r g is r g i ageHY

fl r g i fl r g i

where the three parameters are analogous to those for the length selectivity function above.

For commercial effort, the catchability was written:

(B.4) ( , , ) ( , ) ( , , )season E CSE E S E seasonq r t i q r i s r i t

with ( , )CSE Eq r i being the absolute catchability given by region and effort type, a relative

selectivity coefficient ( , , )S E seasons r i t by region and effort type describing the seasonality of

catchability. Charter boat and recreational fisheries involve non-regional absolute

Page 83: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

74

catchability and relative seasonal catchability parameters. For some regions and

commercial effort types ( , )CSE Eq r i was replaced for limited time periods with an alternative

catchability parameter (separate and specific to each time period).

The instantaneous fishing mortality rate for each element of the population array is

given by a sum of fishing mortalities over all fishing effort types:

(B.5) 1

( , , , ) ( , , , , )E

E

E

n

i

F t r c s F t r c s i

.

The Baranov depletion equation for each element of the population array was written:

(B.6) ( 1, , , ) ( , , , ) exp ( , , , ) ( )yrN t r c s N t r c s M F t r c s p t

where ( )yrp t quantifies the proportion of a year spanned by the days in each half-yearly

time step. Instantaneous natural mortality rate was taken as constant, M = 0.05 yr-1 (Gilbert

et al. 2006; Gilbert, pers. comm.).

6.2.5 Estimation: Parameters and model likelihood

The model likelihood (Fournier and Archibald 1982) was fitted to (1) half-yearly catch totals

by weight (commercial fishery), (2) market sample catch proportions by age, and (3) market

sample catch moment properties of fish length for each age.

Parameters

Estimated parameters for the model fall into six categories: (1) yearly recruit numbers by

region, (2) catchabilities by region (for commercial effort type) and effort type, (3) relative

seasonal selectivities by region (for commercial effort type) and effort type, (4) logistic length

and age selectivity, (5) growth, and (6) likelihood standard deviations of fits to half-yearly

catch totals.

Likelihood for catch totals by weight

Model commercial catch totals by weight (kg) were fitted to data using a lognormal

likelihood. The catch by weight was calculated using the standard Baranov formula as:

(B.7)

ˆ ( , , , , ) ( , , , ) ( ( , ), )

( , , , , ) 1 exp ( , , , ) ( )

( , , , )

E

Eyr

C t r c s i N t r c s w a t c s

F t r c s iM F t r c s p t

M F t r c s

Page 84: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

75

where the weights by age and slice ( ( , ), )w a t c s are derived in Appendix C of Steer et al.

(2016).

The likelihood factor for each combination of region, r, and commercial effort type (iE

up to 2En ), was written:

(B.8)

2

2

1 1 1

ˆln ( , , ) ln ( , , )1 1exp

2 ( , , )2 ( , , ) ( , , )

t E r

E

E E

C

C season EC season E E

nn n

t i r

C t r i C t r iL

t r it r i C t r i

where

( , , )C season Et r i = an estimated catch-likelihood standard deviation parameter, one per effort

type, season, and region;

( , , )EC t r i = reported catch by weight for each time step, t, region, r, and effort type, iE;

ˆ ( , , )EC t r i = model-predicted catch by weight for each t, r, and iE.

A weighted sum-of-squares was used to fit to charter boat ( 1E Ei n ) catches in number

since summer 2007, and similarly for recreational ( E Ei n ) catches in number are fit to for

three telephone and diary surveys run in 2001/02, 2007/08 and 2013/14. Recreational effort

data between the survey years were assumed to vary linearly (by region and season)

between the survey-estimated values and retained the values of 2001/02 prior to that first

survey except for a human population scaling factor, and the values of 2013/14 were

retained post 2013/14. Only catch number data for the full year by region were available

from the 2013/14 survey recently completed, for which we assumed (1) survey catches by

year were broken down into two half-years using proportions taken in winter, and (2) effort

varied relative to 2007/08 values with the same slopes by gulf and season observed for

catches.

Likelihood for catch samples by age

A multinomial likelihood was used to fit to catch-sample proportions by age. The source

data, from the samples per principal gears handline and longline in the half-yearly time

steps and four regions where catch was monitored, consists of the observed counts of

sampled fish falling into each half-yearly age, ( ; )An a i . But the data fitted consists of the

Page 85: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

76

observed counts multiplied by a factor that depends on the relative discrepancy ratio of

each age sampled length value compared to that length in the full market samples of

lengths (including fish not aged), the latter samples taken as being more length-

representative of the population than the aged samples (see the FRDC report, McGarvey

and Feenstra (2004)). Finally, each such corrected count at age-length was multiplied by a

scaling factor so that the total raw sample size is preserved at the level of region, time step,

and gear. The multinomial likelihood factor is written

(B.9) )

48( ;

1

ˆ( ; ) cor

b

A

A

A

n

Ai

n a i

Aa a

iL p a

where

Ai = index over the set of An catch samples of fish ages over half-year, region, and

gear;

)ˆ( ; Ap a i = an array of model-predicted fish proportions captured by age for each

sample indexed by Ai ;

)( ;cor An a i = scaled and corrected observed fish numbers for each age in the catch-

at-age sample Ai .

Likelihood for catch samples by length

A normal likelihood was applied to fit the model to data moment ‘properties’, mean length,

standard deviation of length, skewness, and kurtosis. Fournier and Doonan (1987) first

proposed fitting to length moments and also fitted a normal likelihood, but to the central

moments rather than moment properties. The likelihood for the length moments fit was

written:

(B.10)

2

4 48

1 1

( ; )

ˆ( , ; ) ( , ; )1exp

2

2

A mp

A mp b

A

mp A mp A

n i mp

mp

i i a a mp

n a i

b i a i b i a i

L

.

where

Page 86: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

77

mp = is the estimated moment-likelihood standard deviation parameter, separately per

season, region, and gear.

( , ; )mp Ab i a i = observed moment, indexed by mpi , per sample and half-yearly age.

ˆ( , ; )mp Ab i a i = model-predicted counterpart to ( , ; )mp Ab i a i .

The observed moments were not calculated using the raw counts of fish per age and length

category, but instead were based on length counts from the aged fish that were corrected

for representative length sampling as noted further above (see the FRDC report, McGarvey

and Feenstra (2004)). We weighted each factor in the log-likelihood by the uncorrected

sample size ( ( ; )An a i ), that is by the actual number of aged fish. Higher moment properties

require more data to be informative. We therefore set criteria for exclusion of smaller catch

sample data sets, Ai , from the mpL likelihood, depending on the moment property fitted.

Thus the number of qualifying data sets, ( )A mpn i , decreased with increasing moment

property mpi . We required at least 8 aged fish for kurtosis, 4 for skewness, 2 for standard

deviation, and 1 for fitting to mean length. Similarly we required 4 model slices for kurtosis,

3 for skewness, 2 for standard deviation, and 1 for fitting mean length.

6.2.6 Model performance indicators

Yearly recruitment numbers

Each yearly recruitment number, by region, is a freely estimated model parameter, as

described in Section 6.2.2.

Annual fishable biomass

The annual biomass indicator is the average of the two half-yearly estimates, which for each

region r and year y commencing in October (1983 to 2014) is computed as follows

(B.11)

(2, )

(1, )

1( , ) ( , )

2

season

season

t y

t t y

B y r B t r

.

The fishable biomass at start of each half-yearly time step t in year y ( (1, )seasont y =

October-March, (2, )seasont y = April-September) is given by

Page 87: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

78

(B.12) 23

1 1

( , ) ( , , , ) ( ( , ), )cohort plus nlegs

c cohort s

B t r N t r c s w a t c s

where, for each time step t ranging from 1 (October 1983 to March 1984) to 64 (April to

September 2014), the biomass sum is over cohorts ( c ranging from 1cohort of 2 years olds

of half-yearly age ab, to 23cohort plus of 23 year olds, i.e. the plus group) and over slice

(the slice index s ranging from 1 to nlegs which is the number of length slices of legal size).

The weights by age and slice ( ( , ), )w a t c s are derived in Appendix C of Steer et al. (2016).

Annual harvest fraction

A yearly harvest fraction is defined as the sum of the model-predicted half-yearly catches

divided by the annual average fishable biomass (defined above), as follows

(B.13)

(2, )

1 (1, )

ˆ ( , , )

( , )( , )

season

season

E

E

t y

E

t t y

n

i

C t r i

H y rB y r

where Ei is the index for effort type ranging from 1 (handline) to E

n (recreational).

Annual egg production

The annual egg production indicator is computed as a proportion of pristine, i.e. as a

proportion of an unfished stock. The estimated total annual egg production (1) assumes a

50:50 sex ratio, (2) includes both legal size and undersize females, and (3) employs a

combined fecundity-maturity-at-length relation of 3

0.00072012 ( )l s . The measure of

pristine egg production is obtained by running the model without estimation (as a projection)

with all effort set to zero. Specifically a single equilibrium projected value of egg production

is computed as the average over the last 10 years of a 134 year model period, where annual

recruitment is set to be fixed at the average over historical estimates from 1982-2009. The

estimated annual egg production for the summer half-year of each year y is given by

(B.14)

23

1 0

( , )

( (1, ), , , )exp 0.5 ( (1, ), , , ) ( (1, )) ( ( ))

2

cohort plus nlegs

seasonseason yr season

c cohort s

Eggs y r

N t y r c sM F t y r c s p t y fm l s

where ( ( ))fm l s is the fecundity-maturity function applied to mid-slice length ( )l s .

Page 88: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

79

6.3. Estimating length-at-age and weight-length for the two newly implemented

Gulf St. Vincent regions

Because the spatial resolution of the model was modified, with Gulf St. Vincent partitioned

into northern and southern regions (NGSV and SGSV), prior estimates of parameters

quantifying two basic body size relationships, length-at-age, and weight-at-length, were re-

estimated with data from each region separately.

Growth, as mean and standard deviation of observed body lengths for every age of a cohort

was estimated from market sampled catch lengths-at-age. A normal distribution of each

cohort’s lengths-at-age prior to reaching legal size is assumed. In addition, the knife-edge

cut-off in sampled lengths below the legal minimum length (LML) of 38 cm was explicitly

accounted for by fitting lengths at age to a likelihood pdf that is truncated at 38 cm, meaning

that the probability of observing a sampled fish below that size is set to zero. Sampled

snapper less than 38 cm were removed from this analysis. This truncation length-at-age

estimation method (adapted from that presented in McGarvey and Fowler 2002 for King

George whiting ) avoids bias arising from truncation of catch samples from a fishery with a

strict minimum legal size.

A normal likelihood, prior to imposing truncation at LML, was fitted to model the

distribution of lengths at each age

2

( )1 1exp

2 ( )2 ( )

i i

ii

l l aL

aa

(1)

where il = length of fish sample i , and ia = age of fish sample i , given in half years

obtained from count of its otolith annuli and an assumed birthdate in January of each (year

class) summer spawning.

The mean length-at-age

0( ) 1 exp2

r

ii

a tl a L K

(2)

Page 89: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

80

was modeled by a von Bertalanffy growth formula, generalized by the inclusion of an

exponent, r. Seasonality in growth was not made explicit due to the small number (2) of

yearly time steps.

The likelihood standard deviation () quantifying the spread of normal lengths-at-age for

each half-yearly age was modeled as an allometric function of mean length:

1

0( ) ( )s

i ia s l a . (3)

This power function for standard deviation in terms of mean length has the desired property

that once growth stops, the standard deviation in lengths-at-age also ceases to change.

The left-truncated normal likelihood, which applies to samples from commercial or

recreational fishers,

2 2

( ) ( )1 1 1 1exp exp ,

( ) 2 ( ) ( ) 2 ( )

0,

i i ii

i i i i iLML

i

l l a l l adl if l LML

L a a a a

if l LML

(4)

(4)

postulates a probability cut-off to zero for landed samples less than LML and a renormalized

probability, integrating to 1, for the range of legal lengths.

Parameters were estimated by minimising the negative sum of log-likelihoods using the

ADMB estimation software:

1

ln( )n

i

i

O L

. (5)

The length-at-age curves with associated 95% confidence intervals obtained from Eq. 3

(their fits to data shown in Figures 6.11a and 6.11b) were taken as inputs into SnapEst.

Subsequently, two key growth parameters ( K and 0s ) were further re-estimated in

SnapEst, integrated with the overall stock assessment estimation. Because of the slice-

Page 90: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

81

partition age and length population breakdown in SnapEst, this re-estimation allows for

further correction of growth bias, notably accounting for the asymmetric nature of fishing

mortality which removes faster growing fish from the population at younger ages, namely at

ages when they reach the legal harvestable size range sooner than slower growing fish

(McGarvey and Feenstra 2007).

Parameters for the two Gulf St. Vincent weight-at-length relationships were also re-

estimated for this assessment. Mean weight versus total length was modeled by an

allometric relationship:

( )i iw l l . (6)

A normal likelihood was again used. The standard deviation ( )w il of the likelihood (i.e. of

the fitted spread of observed weights about the mean ( )iw l ) was assumed to vary in a

power relationship with model predicted weight at each given fitted total fish length applying

an analogous error structure to that assumed for length-at-age in Eq. 3:

1

0( ) ( ) w

w i w il w l

. (7)

The resulting weight-length curves (Figures 6.11c and 6.11d) were obtained by minimizing

the negative log-likelihood function (Eq. 5).

Page 91: Snapper (Chrysophrys auratus) Fishery · Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery i Snapper (Chrysophrys auratus) Fishery AJ Fowler, R McGarvey, J Carroll,

Fowler, A. J. et al. (2016) Snapper (Chrysophrys auratus) Fishery

82

Fig. 6.11 Fitted ogives of growth as length-at-age (graphs a and b, for Northern Gulf St. Vincent and Southern Gulf St. Vincent regions respectively) and weight-at-length (c and d). See text for details. Solid blue lines plot the fitted means (Eq. 2 for a and b, Eq. 6 for c and d). Blue error bars show estimated 95% confidence intervals

(1.96 times estimated ( )ia for a and b, and 1.96 times ( )w il for c and d) surrounding all fitted data points

shown.

0 200 400 600 800 1000 1200 1400

Whole

weig

ht (g

)

0

5000

10000

15000

20000c. NGSV length-weight

Total length (mm)

0 200 400 600 800 1000 1200 1400

Whole

weig

ht (g

)

0

5000

10000

15000

20000d. SGSV length-weight

0 5 10 15 20 25 30 35

Tota

l le

ng

th (

mm

)

0

200

400

600

800

1000

1200

1400a. NGSV age-length

Age (years)

0 5 10 15 20 25 30 35

Tota

l le

ng

th (

mm

)

0

200

400

600

800

1000

1200

1400

1600b. SGSV age-length